CN108152458A - Gas detection method and device - Google Patents
Gas detection method and device Download PDFInfo
- Publication number
- CN108152458A CN108152458A CN201711436911.9A CN201711436911A CN108152458A CN 108152458 A CN108152458 A CN 108152458A CN 201711436911 A CN201711436911 A CN 201711436911A CN 108152458 A CN108152458 A CN 108152458A
- Authority
- CN
- China
- Prior art keywords
- data
- gas
- content
- parameter
- environment
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 106
- 238000012360 testing method Methods 0.000 claims abstract description 166
- 238000012549 training Methods 0.000 claims abstract description 76
- 238000012544 monitoring process Methods 0.000 claims abstract description 43
- 238000000034 method Methods 0.000 claims abstract description 36
- 230000002068 genetic effect Effects 0.000 claims description 68
- 238000004422 calculation algorithm Methods 0.000 claims description 19
- 230000001186 cumulative effect Effects 0.000 claims description 15
- 230000006978 adaptation Effects 0.000 claims description 12
- 230000000052 comparative effect Effects 0.000 claims description 10
- 108010001267 Protein Subunits Proteins 0.000 claims description 3
- 238000010276 construction Methods 0.000 claims description 3
- 239000007789 gas Substances 0.000 description 284
- 230000007613 environmental effect Effects 0.000 description 29
- 230000006870 function Effects 0.000 description 28
- 238000004364 calculation method Methods 0.000 description 24
- 230000008569 process Effects 0.000 description 12
- UGFAIRIUMAVXCW-UHFFFAOYSA-N Carbon monoxide Chemical compound [O+]#[C-] UGFAIRIUMAVXCW-UHFFFAOYSA-N 0.000 description 7
- 229910002091 carbon monoxide Inorganic materials 0.000 description 7
- 241000208340 Araliaceae Species 0.000 description 6
- WSFSSNUMVMOOMR-UHFFFAOYSA-N Formaldehyde Chemical compound O=C WSFSSNUMVMOOMR-UHFFFAOYSA-N 0.000 description 6
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 6
- 235000003140 Panax quinquefolius Nutrition 0.000 description 6
- 235000008434 ginseng Nutrition 0.000 description 6
- 230000008878 coupling Effects 0.000 description 4
- 238000010168 coupling process Methods 0.000 description 4
- 238000005859 coupling reaction Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 238000013178 mathematical model Methods 0.000 description 4
- 230000035772 mutation Effects 0.000 description 4
- 238000003860 storage Methods 0.000 description 4
- 230000010354 integration Effects 0.000 description 3
- 238000005457 optimization Methods 0.000 description 3
- 239000002341 toxic gas Substances 0.000 description 3
- 238000003491 array Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 239000003034 coal gas Substances 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000008303 genetic mechanism Effects 0.000 description 2
- 230000000977 initiatory effect Effects 0.000 description 2
- 238000007689 inspection Methods 0.000 description 2
- 238000009434 installation Methods 0.000 description 2
- 238000012804 iterative process Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 238000012706 support-vector machine Methods 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 1
- 238000011109 contamination Methods 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 230000007717 exclusion Effects 0.000 description 1
- 230000014759 maintenance of location Effects 0.000 description 1
- 238000012821 model calculation Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000000505 pernicious effect Effects 0.000 description 1
- 231100000614 poison Toxicity 0.000 description 1
- 230000007096 poisonous effect Effects 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
- G01N33/0004—Gaseous mixtures, e.g. polluted air
- G01N33/0009—General constructional details of gas analysers, e.g. portable test equipment
- G01N33/0062—General constructional details of gas analysers, e.g. portable test equipment concerning the measuring method or the display, e.g. intermittent measurement or digital display
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
- G01N33/0004—Gaseous mixtures, e.g. polluted air
- G01N33/0009—General constructional details of gas analysers, e.g. portable test equipment
- G01N33/0062—General constructional details of gas analysers, e.g. portable test equipment concerning the measuring method or the display, e.g. intermittent measurement or digital display
- G01N33/0068—General constructional details of gas analysers, e.g. portable test equipment concerning the measuring method or the display, e.g. intermittent measurement or digital display using a computer specifically programmed
Landscapes
- Chemical & Material Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Medicinal Chemistry (AREA)
- Food Science & Technology (AREA)
- Combustion & Propulsion (AREA)
- Physics & Mathematics (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Investigating Or Analyzing Materials By The Use Of Fluid Adsorption Or Reactions (AREA)
Abstract
This application discloses a kind of gas detection method and device, the method includes：It determines to acquire the content detection data of the gas to be monitored of acquisition and the environment measuring data of at least one envirment factor in current environment；By the content detection data and at least one environment measuring data input gas content computation model that training obtains in advance, the content monitoring data for obtaining the gas to be monitored in current environment is calculated；Wherein, the content monitoring data is used to assess whether the gas to be monitored pollutes current environment；Content preset data of the gas content computation model based on the gas to be monitored preset in test environment, the environment preset data of at least one envirment factor and the test data training acquisition in advance obtained in the test environment by sensor array acquisition.The present invention improves the accuracy of detection of gas to be monitored.
Description
Technical field
The application belongs to field of intelligent control technology, specifically, being related to a kind of gas detection method and device.
Background technology
Coal gas is a kind of energy of clean and effective, is used widely in family of China.Due to main component in coal gas
For pernicious gases such as carbon monoxide, during improper use, easily threaten to the person.It is often necessary to being equipped with gas piping
Room in install gas monitor equipment, by the gas monitor monitoring of tools whether there is corresponding gas pollution.
In the prior art, containing for the carbon monoxide in air namely in environment is typically acquired using carbon monoxide transducer
Amount, the carbon monoxide content obtained based on acquisition judge whether gas pollution.
But since the usual antijamming capability of gas sensor is poor, the carbon monoxide in the air of detection etc. is caused to have
The content of poisonous gas is inaccurate.
Invention content
In view of this, this application provides a kind of gas detection method and devices, to solve in the prior art, traditional gas
Sensor anti-interference is poor, the technical issues of causing the toxic gases content detection such as CO in Air inaccurate.
In order to solve the above-mentioned technical problem, the application first aspect provides a kind of gas detection method, including：
Determine in current environment by sensor array acquisition obtain gas to be monitored content detection data and
The environment measuring data of at least one envirment factor；By the content detection data and at least one environment measuring data
The input gas content computation model that training obtains in advance calculates the content prison for obtaining the gas to be monitored in current environment
Control data；Wherein, the content monitoring data is used to assess whether the gas to be monitored pollutes current environment；The gas contains
Measure content preset data of the computation model based on the gas to be monitored preset in test environment, at least one environment because
The environment preset data of son and the test data obtained in the test environment by sensor array acquisition are trained in advance
It obtains.
Preferably, specifically training obtains the gas content computation model in advance as follows：
Determine the corresponding at least one set of training data of at least one test environment；Wherein, each group of training data is included in
The ring of the content preset data of the preset gas to be monitored, at least one envirment factor in its corresponding test environment
Border preset data and the test data obtained in the test environment by sensor array acquisition；Build gas content
Computation model；Based at least one set of training data, training obtains the model parameter of the gas content computation model.
Preferably, described based at least one set of training data, training obtains the mould of the gas content computation model
Shape parameter includes：
Determine at least one set of reference parameter of the gas content computation model；By the content preset data and described
In the inversion model of the corresponding gas content computation model of the environment preset data input at least one set reference parameter, obtain
The corresponding at least one prediction data of institute's test environment；At least one prediction data is carried out respectively with the test data
Compare；Based on comparative result, optimum prediction data are selected from least one prediction data；Judge the optimum prediction number
According to whether meeting required precision；If so, determine the corresponding one group of object reference parameter of the optimum prediction data；By the mesh
Mark model parameter of the reference parameter as the gas content computation model；If not, based on the object reference parameter, obtain
At least one set of genetic parameter；Using at least one set of genetic parameter as at least one set of reference parameter, return to described
The step of at least one set of reference parameter for determining the gas content computation model, continues to execute.
Preferably, it is described using at least one set of genetic parameter as at least one set of reference parameter, return to institute
After the step of at least one set of reference parameter for stating the determining gas content computation model continues to execute, the method is also wrapped
It includes：
Record performs cumulative number；It is described to be based on the object reference parameter, obtain at least one set of genetic parameter；By described in
At least one set of genetic parameter returns to the determining gas content computation model as at least one set of reference parameter
At least one set of reference parameter the step of continue to execute including：
Judge whether current execution cumulative number is more than maximum and performs number；
If so, determine the corresponding one group of object reference parameter of the optimum prediction data；By the object reference parameter
Model parameter as the gas content computation model；
If not, based on the object reference parameter, at least one set of genetic parameter is obtained；By at least one set of heredity ginseng
Number returns at least one set of reference for determining the gas content computation model as at least one set of reference parameter
The step of parameter, continues to execute.
Preferably, it is described at least one prediction data is compared respectively with the test data including：
Based on fitness function, each prediction data and the test data are inputted into the fitness function respectively
It is calculated, obtains at least one fitness value；
It is described based on comparative result, from least one prediction data select optimum prediction data include：
Based at least one fitness value, the maximum adaptation angle value at least one fitness value is determined；
It is the optimum prediction data to determine the corresponding prediction data of the maximum adaptation degree.
Preferably, it is described based on the object reference parameter, it obtains at least one set of genetic parameter and includes：
Using the object reference parameter as genetic operator, at least one set of genetic parameter is obtained using genetic algorithm.
Preferably, at least one envirment factor includes：Environment temperature, ambient humidity and interference gas；
The content detection data of the gas to be monitored of acquisition and at least one environment are acquired in the determining current environment
The environment measuring data of the factor include：
Determine to acquire the content detection data of the gas to be monitored of acquisition and the environment temperature in current environment
The interference content data of temperature detection data, the Humidity Detection data of the ambient humidity and the interference gas.
The embodiment of the present invention also provides a kind of gas-detecting device, including：
First determining module, for determine the content detection data for the gas to be monitored for being acquired in current environment acquisition with
And the environment measuring data of at least one envirment factor；
First computing module, it is pre- for the content detection data and at least one environment measuring data to be inputted
The gas content computation model that first training obtains calculates the content monitoring number for obtaining the gas to be monitored in current environment
According to；
Wherein, the content monitoring data is used to assess whether the gas to be monitored pollutes current environment；The gas
Content preset data of the content calculation model based on the gas to be monitored preset in test environment, at least one environment
The environment preset data of the factor and the test data obtained in the test environment by sensor array acquisition are instructed in advance
Practice and obtain.
Preferably, described device is especially by with lower module, training obtains gas content computation model in advance：
Second determining module, for determining the corresponding at least one set of training data of at least one test environment；Wherein, it is each
Group training data be included in the content preset data of the preset gas to be monitored in its corresponding test environment, it is described at least
The environment preset data of one envirment factor and pass through the test number that sensor array acquisition obtains in the test environment
According to；
Model construction module, for building gas content computation model；
Parameter training module, for being based at least one set of training data, training obtains the gas content and calculates mould
The model parameter of type.
Preferably, the parameter training module includes：
Reference parameter unit, for determining at least one set of reference parameter of the gas content computation model；
Content prediction unit, for will the content preset data and the environment preset data input described at least one
In the inversion model of the corresponding gas content computation model of group reference parameter, it is corresponding at least one pre- to obtain institute's test environment
Measured data；
First comparing unit, at least one prediction data to be compared respectively with the test data；
First selecting unit, for based on comparative result, optimum prediction number being selected from least one prediction data
According to；
First judging unit, for judging whether the optimum prediction data meet required precision；
First determination unit, for if so, determining the corresponding one group of object reference parameter of the optimum prediction data；It will
Model parameter of the object reference parameter as the gas content computation model；
Second determination unit, for if not, based on the object reference parameter, obtaining at least one set of genetic parameter；It will
At least one set genetic parameter returns to the determining gas content and calculates as at least one set of reference parameter
The step of at least one set of reference parameter of model, continues to execute.
Preferably, the parameter training model further includes：Number accumulated unit performs cumulative number for recording；It is described
Second determination unit includes：Judgment sub-unit, for judging currently to perform whether cumulative number is more than maximum execution number；First
Determination subelement, for if so, determining the corresponding one group of object reference parameter of the optimum prediction data；The target is joined
Examine model parameter of the parameter as the gas content computation model；Second determination subelement, for if not, based on the mesh
Reference parameter is marked, obtains at least one set of genetic parameter；Using at least one set of genetic parameter as described at least one set of with reference to ginseng
The step of number, at least one set of reference parameter for returning to the determining gas content computation model, continues to execute.
Preferably, first comparing unit includes：
Fitness computation subunit, for being based on fitness function, respectively by each prediction data and the test number
It is calculated according to the fitness function is inputted, obtains at least one fitness value；
The first selecting unit includes：
Optimal selection subelement for being based at least one fitness value, determines at least one fitness value
In maximum adaptation angle value；
Target determination subelement, for determining that the corresponding prediction data of the maximum adaptation degree is the optimum prediction number
According to.
Preferably, second determination unit includes：
Third determination subelement, for using the object reference parameter as genetic operator, institute to be obtained using genetic algorithm
State at least one set of genetic parameter.
Preferably, at least one envirment factor includes：Environment temperature, ambient humidity and interference gas；Described
One determining module includes：
Data determination unit, for determine the content detection data for the gas to be monitored for being acquired in current environment acquisition with
And the interference of the temperature detection data of the environment temperature, the Humidity Detection data of the ambient humidity and the interference gas
Content data.
In the embodiment of the present invention, it may be determined that the gas to be monitored obtained in current environment by sensor array acquisition
The environment measuring data of content detection data and at least one envirment factor, not just for be detected in air during monitoring
Gas content data are monitored also directed to other envirment factors in air, and the content of monitoring is more comprehensive.It is to be monitored when obtaining
During the environment measuring data of the content detection data of gas and at least one envirment factor, training in advance can be inputted
Gas content computation model can be calculated the content obtained in gas to be monitored by the gas content computation model and monitor number
According to the content monitoring data is whether the gas to be monitored pollutes current environment.By the way that environmental factor is increased to ring
Gas content computation model is used so that result of calculation is more accurate, therefore obtain during border considers and in calculating process
The gas testing result to be monitored obtained is also more accurate.
Description of the drawings
Attached drawing described herein is used for providing further understanding of the present application, forms the part of the application, this Shen
Illustrative embodiments and their description please do not form the improper restriction to the application for explaining the application.In the accompanying drawings：
Fig. 1 is a kind of flow chart of one embodiment of gas detection method of the embodiment of the present application；
Fig. 2 is the flow chart of one embodiment of the gas content computation model training process of the embodiment of the present application；
Fig. 3 is the flow chart of another embodiment of the gas content computation model training process of the embodiment of the present application；
Fig. 4 is the flow chart of another embodiment of the gas content computation model training process of the embodiment of the present application；
Fig. 5 is the flow chart of one embodiment of the gas detection method of the embodiment of the present application；
Fig. 6 is a kind of structure diagram of one embodiment of gas-detecting device of the embodiment of the present invention；
Fig. 7 is the structure diagram of one embodiment of the gas content computation model training device of the embodiment of the present application.
Specific embodiment
Presently filed embodiment is described in detail below in conjunction with accompanying drawings and embodiments, thereby how the application is applied
Technological means can fully understand and implement according to this to solve technical problem and reach the realization process of technical effect.
The embodiment of the present invention mainly apply in environmental monitoring, by by the increase of envirment factor and gas content meter
The work such as increase of model are calculated, the accuracy of detection of gas to be monitored in environment can be improved.
In the prior art, usually using single gas sensor, for example, electrochemical gas sensor, semi-conductor gas
Sensor etc. is monitored the gas to be monitored in air, obtains the aerial content data of gas to be monitored.But this
Kind gas sensor carries out single monitoring only for gas to be monitored, passes through the safety that corresponding monitoring function realizes environment
Assessment.But different environmental factors may be included in environment, for example, temperature, humidity and other kinds of gas content
Deng, and then, the monitoring only for single kind can cause the inaccuracy of monitored results, it is impossible to accurate judgement current environment whether by
To pollution.
Therefore, inventor expects whether to acquire the content of various environmental factors, to reduce other factors pair in environment
Influence caused by monitoring process.It when acquiring various environmental factors, needs to acquire the data of environment in a variety of environment, also just needs
Multiple sensors are wanted, in order to obtain high-precision sensing data, inventor expects using by a plurality of types of sensor integrations
MEMS (Micro-Electro-Mechanical System, microelectromechanical systems) sensor array, passes through sensor array
Acquire the content detection data of gas to be monitored and the environment measuring data of at least one envirment factor.Detecting environment inspection
After measured data, if with the ratio of simple concentration and total amount, calculating process is not very accurate, be easy to cause monitored results
Inaccuracy, therefore, inventor expect whether mathematics computing model can be utilized, and calculate the content monitoring number of the gas to be monitored
According to.Accordingly, the technical solution of the application is inventors herein proposed.
In the embodiment of the present invention, by the content detection data and at least for acquiring the gas to be monitored in current environment
The content detection data and the environment measuring data are input to gas by the environment measuring data of one envirment factor later
Body content calculation model, wherein, the gas content computation model is according to the gas to be monitored preset in test environment
Content preset data, the environment preset data of at least one envirment factor and pass through biography in the test environment
The test data that sensor array acquisition obtains training in advance obtains, can by the calculating of above-mentioned gas content model calculation formula
To obtain accurate gas content data.It is monitored thus, it is possible to calculate the content obtained by above-mentioned gas content calculation model
Data judge whether there is the gaseous contamination to be monitored in current environment.Improve the accuracy of detection.
The embodiment of the present invention is described in detail below in conjunction with attached drawing.
As shown in Figure 1, for a kind of flow chart of one embodiment of gas detection method provided in an embodiment of the present invention, institute
The method of stating can include following steps：
101：It determines to acquire the content detection data of the gas to be monitored of acquisition and at least one ring in current environment
The environment measuring data of the border factor.
102：By the content detection data and at least one environment measuring data input gas trained in advance
Content calculation model calculates the content monitoring data for obtaining the gas to be monitored in current environment.
Wherein, the content monitoring data is used to evaluate whether the gas to be monitored pollutes current environment；The gas
Content preset data of the content calculation model based on the gas to be monitored preset in test environment, at least one environment
The environment preset data of the factor and the test data obtained in the test environment by sensor array acquisition are instructed in advance
Practice and obtain.
The gas to be monitored as needs the toxic gas monitored, for example, the gases such as carbon monoxide, formaldehyde.The biography
Sensor array can include MEMS sensor array, and the acquisition obtains the content detection data of gas to be monitored in current environment
And the environment measuring data of at least one envirment factor can obtain current environment using MEMS sensor array acquisition
In gas to be monitored content detection data and be environment measuring data in an envirment factor.
The content detection data of the gas to be monitored be MEMS sensor array acquisition to current environment in wait to supervise
Control the related data of the content of gas.The data class of the content detection data is not defined herein, can be with height
The high resistant data that resistance form represents.The environment of the gas content detection data to be monitored and at least one envirment factor
Detection data be sensor array acquisition obtain with above-mentioned gas content and the data of at least one envirment factor
And corresponding value of electrical signals.
Described in sensor that can be including the detection gas to be monitored in the MEMS sensor array and detection extremely
The sensor of a few envirment factor, can be made by different types of sensor integration in a sensor array with facilitating
With.
The gas content computation model, can refer to can be used for according to content detection data and environment measuring data with
Calculate the mathematical model of gas content.It can include different parameters, the computation model in the gas content computation model
Calculating process mainly pass through the different parameter and calculate and obtain.
The gas content computation model need using be obtained ahead of time content preset data, environment preset data and
The test data obtained in the test environment by sensor array acquisition, which is trained, to be obtained.The content preset data and
The environment preset data is setting data, is mainly built according to the content preset data and the environment preset data
Test environment.For example, when the content preset data is 15%, that is, the content of the gas to be monitored in a test environment is set
It is set as 15%.Optionally, the test data can include the content measuring data and at least one of the gas to be monitored
A envirment factor test data.The test data refers to the data obtained using MEMS sensor array test, using described
MEMS sensor array can detect the content measuring data of gas to be monitored in preset test environment and the test
The environmental testing data of at least one envirment factor in environment.
Optionally, it is described by the content detection data and at least one environment measuring data input training in advance
Gas content computation model before, the method can also include：By content detection data and described at least one
Environment measuring data are normalized.
It is described by the content detection data and gas that at least one environment measuring data input is trained in advance
Content calculation model can include：By the content detection data after the normalized and at least one environment inspection
Measured data input gas content computation model trained in advance.
By the way that the normalized of the content detection data and at least one environment measuring data can be made
Two kinds of data belong to same type of data, and there is no the situations of data class or Type-Inconsistencies, more convenient when calculating.
In the embodiment of the present invention, in the content detection data of gas to be monitored in monitoring current environment, also directed to environment
In other envirment factors be detected, obtain the environment measuring data of at least one envirment factor, the content detection data with
And the environment measuring data be input to the gas content computation model after can obtain gas to be monitored in current environment
In content monitoring data.The content monitoring data can improve the accuracy of monitoring due to consideration that various environmental factors.
As one embodiment, at least one envirment factor can include：Environment temperature, ambient humidity and interference
Gas；
The content detection data of the gas to be monitored of acquisition and at least one environment are acquired in the determining current environment
The environment measuring data of the factor can include：
Determine to acquire the content detection data of the gas to be monitored of acquisition and the environment temperature in current environment
The interference content data of temperature detection data, the Humidity Detection data of the ambient humidity and the interference gas.
The content detection data for determining to acquire the gas to be monitored of acquisition in current environment and the environment temperature
The interference content data of the temperature detection data of degree, the Humidity Detection data of the ambient humidity and the interference gas can be with
Including：
Determine in current environment using first gas sensor acquisition obtain gas to be monitored content detection data,
The Humidity Detection data and second gas that the temperature detection data of temperature sensor acquisition acquisition, humidity sensor acquisition obtain
The interference content data that sensor acquisition obtains.
The first gas sensor, temperature sensor, humidity sensor and second gas sensor may be used
Above-mentioned MEMS sensor can be specifically integrated into an array by MEMS sensor, and form MEMS sensor array is made with facilitating
With.
In the embodiment of the present invention, by different types of envirment factors such as temperature, humidity, the interference gas in environment
Under effect, the content for obtaining corresponding gas to be monitored can be calculated, considering the influence of the multiple environment factor can accurately define
The influence of different factor pairs gas to be monitored, obtains more accurate data influence result in environment.
As another embodiment, as shown in Fig. 2, for the gas content computation model training process in the embodiment of the present application
One embodiment flow chart, the model specifically can according to following steps in advance training obtain：
201：Determine the corresponding at least one set of training data of at least one test environment.
Wherein, each group of training data is included in the content of the preset gas to be monitored in its corresponding test environment
Preset data, the environment preset data of at least one envirment factor and pass through sensor array in the test environment
The test data that row acquisition obtains.
202：Build gas content computation model；
203：Based at least one set of training data, training obtains the model parameter of the gas content computation model.
At least one test environment refers to, pre-set the content of each gas and the numerical value of envirment factor with
For the test environment of test.The corresponding content preset data of the test environment is the content of preset gas to be monitored
Data, environment preset data are the data of other preset envirment factors.At least one test environment is according to
Content preset data and environment preset data and build what is finished.At least one envirment factor can be in above-described embodiment
Described environment temperature, ambient humidity and interference gas etc., when setting test environment, can set respectively environment temperature,
The content of ambient humidity, interference gas and gas to be monitored, to build corresponding test environment.
By MEMS sensor array when detection devices are placed in each test environment, the MEMS sensor array
Wait detection devices that can test the content for monitoring gas and at least one environment in the environment in each test environment
The numerical value of the factor namely the test data that the test environment can be determined by sensor arrays such as MEMS, obtain various gas
The test data of body and envirment factor.The test data can include the content measuring data and at least of gas to be monitored
The environmental testing data of one envirment factor.
The gas content computation model refers to the mathematical model that can be used for calculating gas content.As a kind of possible
Realization method, the gas content computation model can refer to LSSVM (Least Squares Support Vector
Machine, least square method supporting vector machine) model, LSSVM models can include multiple parameters.The structure gas content meter
The multiple parameter values that model specifically refers to determine the LSSVM models are calculated, determine to make by the multiple parameter values for determining to use
The calculation formula of LSSVM at this point, determining the multiple parameter values used as unknown number, needs to utilize at least one set of training number
According to training obtains the model parameter of the gas content computation model.
As a kind of possible realization method, as shown in figure 3, above-mentioned steps 203：It is described to be trained based on at least one set
Model, the model parameter that training obtains the gas content computation model can be realized according to following steps：
301：Determine at least one set of reference parameter of the gas content computation model.
302：By the content preset data and environment preset data input at least one set of reference parameter point
In the inversion model of not corresponding gas content computation model, the corresponding at least one prediction data of institute's test environment is obtained.
Wherein, at least one preset data can refer to is obtained by the inversion model of gas content computation model by calculating
The content prediction data of gas to be monitored and the environmental forecasting number of at least one envirment factor in the test environment obtained
According to.
303：At least one prediction data is compared respectively with the test data.
The test data can include the content measuring data of gas to be monitored obtained by sensor array acquisition
And the test data of at least one envirment factor.It is described by least one prediction data respectively with the test data into
Row can relatively include：
The content prediction data of gas to be monitored with content measuring data are compared, obtain gas comparison result；
The environmental forecasting data of each envirment factor with environmental testing data are compared respectively, obtain environment
Comparison result.
More targetedly comparison result can be obtained by being respectively compared, for respectively from gas and envirment factor
Angle weighs the difference of test data and prediction data.
It is described that at least one prediction data is compared with the test data can also include respectively：
By the content prediction data in prediction data and at least one environmental forecasting data, with the content in test data
Test data and at least one environmental testing data, are carried out at the same time and compare, and obtain whole comparison result.
Pass through the whole whole comparison result that can relatively obtain prediction data and test data so that comparison result has more
Whole difference.
304：Based on comparative result, optimum prediction data are selected from least one prediction data.
305：Judge whether the optimum prediction data meet required precision.
306：If so, determine the corresponding one group of object reference parameter of the optimum prediction data；By the object reference
Model parameter of the parameter as the gas content computation model.
307：If not, based on the object reference parameter, at least one set of genetic parameter is obtained；By at least one set of something lost
Parameter is passed as at least one set of reference parameter, returns at least one set for determining the gas content computation model
The step of reference parameter, continues to execute.
Optionally, when the gas content computation model refers specifically to the LSSVM models, the model parameter of the LSSVM
Penalty factor and kernel function can specifically be referred to, the penalty factor and kernel function can be used as described in the embodiment of the present application
Reference parameter.
Determining at least one set of reference parameter can refer to, and according to needs are used, the parameter is carried out Initialize installation.
When the gas content computation model refers specifically to the LSSVM models, set at least one set of reference parameter that can specifically refer to described
The initialization data of penalty factor and kernel function, respectively described penalty factor and kernel function set an initialization data
When, you can to determine one group of initiation parameter.
The gas content computation model refers to what is obtained using sensor array acquisition, with the gas content to be monitored
The corresponding electric signal output value of the environment measuring data of detection data and at least one envirment factor, by the telecommunications
Number output valve can obtain the content monitoring data of the gas to be monitored and the environmental monitoring number of envirment factor by calculating
According to computation model.
Similarly, the environment of the content preset data of the gas to be monitored and at least one envirment factor is preset
Data when inputting the inversion model of the gas content computation model, can calculate acquisition and contain with what the sensor array acquired
The electric signal output for measuring test data and environmental testing data is worth corresponding content prediction data and environmental forecasting data.
It is described to judge whether the optimum prediction data meet required precision and refer to, by the optimum prediction data and survey
Data input precision calculation formula is tried, judges whether the precision difference of the two meets preset condition, is preset for example, mean square deviation is less than
Threshold value when meeting, that is, meets required precision.
By setting at least one set of reference parameter, using the reference parameter as training basis, based on the gas content
The inversion model of computation model carries out model training, screens legal optimum prediction data, expires in the optimum prediction data
During sufficient required precision, you can to determine corresponding object reference parameter, to determine corresponding object reference model as model parameter.
If the optimum prediction data are unsatisfactory for required precision, object reference parameter can be directed to as determining corresponding at least one
Group genetic parameter redefines corresponding model parameter.The comparison of prediction data and test data may be constructed the of model parameter
One layer of selection criteria, the second layer selection criteria for being selected as model parameter selection of required precision, bilayer selection may insure
Select corresponding more accurate model parameter.
In certain embodiments, it is described using at least one set of genetic parameter as at least one set of reference parameter, weight
After the step of at least one set of reference parameter for newly returning to the determining gas content computation model, continues to execute, the side
Method can also include：
Record performs cumulative number；
It is described to be based on the object reference parameter, obtain at least one set of genetic parameter；By at least one set of genetic parameter
As at least one set of reference parameter, at least one set of reference ginseng for determining the gas content computation model is returned to
Several steps continue to execute including：
Judge whether current execution cumulative number is more than maximum and performs number；
If so, determine the corresponding one group of object reference parameter of the optimum prediction data；By the object reference parameter
Model parameter as the gas content computation model；
If not, based on the object reference parameter, at least one set of genetic parameter is obtained；By at least one set of heredity ginseng
Number returns at least one set of reference for determining the gas content computation model as at least one set of reference parameter
The step of parameter, continues to execute.
As shown in figure 4, to be added to cumulative number judgment rule step 401：Judge whether current execution cumulative number is big
In the flow chart of one embodiment of the training process of the maximum above-mentioned model parameter for performing number.
In above-described embodiment, when optimum prediction data are unsatisfactory for required precision, need to carry out model parameter hereditary excellent
Change and calculate.But genetic optimization calculating is an iterative calculation, limit perform number can to avoid occurring iterative process always,
Iterative calculation performs and can not obtain model parameter always.It can be with the continuity of guarantee procedure.
In certain embodiments, described be compared at least one prediction data with the test data respectively can
To include：
Based on fitness function, each prediction data and the test data are inputted into the fitness function respectively
It is calculated, obtains at least one fitness value；
It is described based on comparative result, from least one prediction data select optimum prediction data can include：
Based at least one fitness value, the maximum adaptation angle value at least one fitness value is determined；
It is the optimum prediction data to determine the corresponding prediction data of the maximum adaptation degree.
Based on fitness function, by least one sensor array prediction data and the Sensor array number
It is calculated according to the fitness function is inputted, to obtain at least one fitness value, by pair of the fitness value maximum of acquisition
The prediction data answered, as the optimum prediction data.
In above-described embodiment, at least one prediction data and test data are evaluated by using fitness function
Difference can improve the computational accuracy of difference by the use of fitness function, to select optimal model parameter, obtain compared with
The monitored results of good gas content computation model, the then final more accurate gas to be monitored of acquisition.
In certain embodiments, described based on the object reference parameter, obtaining at least one set of Mutation parameter can include：
Using the object reference parameter as genetic operator, at least one set of genetic parameter is obtained using genetic algorithm.
It is described using the object reference parameter as genetic operator, using in genetic algorithm selection operation, crossover operation,
After mutation operation, more newly-generated a new generation population namely at least one set of genetic parameter, the genetic parameter may act as newly
At least one reference parameter of a generation.
The genetic algorithm can specifically refer to (Genetic Algorithm, GA algorithm), refer specifically to model Darwin biology
The computation model of the biological evolution process of evolutional natural selection and genetic mechanisms.The computation model combination above-described embodiment
Described in LSSVM algorithms when, each step of above-described embodiment still can be utilized to carry out the model parameter of gas to be monitored
It calculates, obtains the gas content computation model that GA-LSSVM is formed.
By reference to the setting of parameter, determining for the gas content computation model can be realized, by by preset gas
Body content preset data and environment preset data are input in the inversion model of corresponding gas content computation model, pass through the gas
After the inversion model of body content calculation model calculates acquisition prediction data, the prediction data is compared with test data, with
It determines optimal reference parameter, and then obtains the object reference parameter.
In the embodiment of the present invention, pass through the corresponding at least one set of test data of preset at least one test environment, training
The gas content computation model, obtains the model parameter of the gas content computation model, and model parameter is based at least one
The initial model parameter training of group obtains, and by repeatedly training, can obtain accurate model parameter so that the gas content
Computation model more levels off to accurate content calculation precision, accordingly, can improve containing for gas to be monitored in the embodiment of the present invention
The result of calculation of monitoring data is measured, to obtain accurate environmental monitoring result.
As shown in figure 5, the flow chart of one embodiment for gas detection method, this method may comprise steps of：
501：Determine the corresponding at least one set of training data of at least one test environment.
Wherein, each group of training data is included in the content of the preset gas to be monitored in its corresponding test environment
Preset data, the environment preset data of at least one envirment factor and pass through sensor array in the test environment
The test data that row acquisition obtains.
502：Build LSSVM gas content computation models.Wherein, the model ginseng of the LSSVM gas contents computation model
Number is penalty factor and kernel function.
503：Determine penalty factor and the corresponding at least one set of reference parameter of kernel function；
504：By the content preset data and environment preset data input at least one set of reference parameter point
In the inversion model of not corresponding LSSVM models, the corresponding prediction data of the test environment is obtained.
505：At least one prediction data is compared respectively with the test data.
506：Based on comparative result, optimum prediction data are selected from least one prediction data.
507：Judge whether the optimum prediction data meet required precision, if step 509 is performed, if not, holding
Row step 508.
508：Judge whether current execution cumulative number is more than maximum and performs number, if so, step 510 is performed, if
It is no, perform step 509.
509：If so, determine the corresponding one group of object reference parameter of the optimum prediction data；By the object reference
Model parameter of the parameter as the gas content computation model obtains GA-LSSVM models.
The LSSVM gas contents computation model can be marked due to carrying out parameter optimization using GA algorithms with GA-LSSVM
Know the model.
510：If not, using the object reference parameter as genetic operator 5101, the selection operation in GA algorithms is carried out
5102nd, after crossover operation 5103, mutation operation 5104, more newly-generated a new generation population 5105 namely at least one set of heredity ginseng
Number；Using at least one set of genetic parameter as at least one set of reference parameter, return to it is described determine the punishment because
The step of son and kernel function corresponding at least one set of reference parameter, continues to execute.
511：It determines to acquire the content detection data of the gas to be monitored of acquisition and at least one ring in current environment
The environment measuring data of the border factor.
512：By the content detection data and at least one environment measuring data input GA- trained in advance
LSSVM models calculate the content monitoring data for obtaining the gas to be monitored in current environment.
In the embodiment of the present invention, gas content computation model of the GA-LSSVM algorithm models to gas to be monitored is first passed through
LSSVM carries out model parameter calculation, calculates after completing, and utilizes the content detection data of detection and described at least one
Environment measuring data input GA-LSSVM models trained in advance, obtain accurate gas content.
As shown in fig. 6, for a kind of structural representation of one embodiment of gas-detecting device provided in an embodiment of the present invention
Figure, described device can include following module：
First determining module 601, for determining the content detection number for the gas to be monitored for being acquired in current environment acquisition
According to this and the environment measuring data of at least one envirment factor；
First computing module 602, for the content detection data and at least one environment measuring data are defeated
Enter the gas content computation model that training obtains in advance, calculate the content monitoring for obtaining the gas to be monitored in current environment
Data；
Wherein, the content monitoring data is used to assess whether the gas to be monitored pollutes current environment；The gas
Content preset data of the content calculation model based on the gas to be monitored preset in test environment, at least one environment
The environment preset data of the factor and the test data obtained in the test environment by sensor array acquisition are instructed in advance
Practice and obtain.
Wherein, the content monitoring data is used to evaluate whether the gas to be monitored pollutes current environment；The gas
Content preset data of the content calculation model based on the gas to be monitored preset in test environment, at least one environment
The environment preset data of the factor and the test data obtained in the test environment by sensor array acquisition are instructed in advance
Practice and obtain.
The gas to be monitored as needs the toxic gas monitored, for example, the gases such as carbon monoxide, formaldehyde.The biography
Sensor array can include MEMS sensor array, and the acquisition obtains the content detection data of gas to be monitored in current environment
And the environment measuring data of at least one envirment factor can obtain current environment using MEMS sensor array acquisition
In gas to be monitored content detection data and be environment measuring data in an envirment factor.
The content detection data of the gas to be monitored be MEMS sensor array acquisition to current environment in wait to supervise
Control the related data of the content of gas.The data class of the content detection data is not defined herein, can be with height
The high resistant data that resistance form represents.The environment of the gas content detection data to be monitored and at least one envirment factor
Detection data be sensor array acquisition obtain with above-mentioned gas content and the data of at least one envirment factor
And corresponding value of electrical signals.
Described in sensor that can be including the detection gas to be monitored in the MEMS sensor array and detection extremely
The sensor of a few envirment factor, can be made by different types of sensor integration in a sensor array with facilitating
With.
The gas content computation model, can refer to can be used for according to content detection data and environment measuring data with
Calculate the mathematical model of gas content.It can include different parameters, the computation model in the gas content computation model
Calculating process mainly pass through the different parameter and calculate and obtain.
The gas content computation model need using be obtained ahead of time content preset data, environment preset data and
The test data obtained in the test environment by sensor array acquisition, which is trained, to be obtained.The content preset data and
The environment preset data is setting data, is mainly built according to the content preset data and the environment preset data
Test environment.For example, when the content preset data is 15%, that is, the content of the gas to be monitored in a test environment is set
It is set as 15%.Optionally, the test data can include content measuring data and at least one envirment factor test number
According to.The test data refers to the data obtained using MEMS sensor array test, can using the MEMS sensor array
To detect at least one ring in the content measuring data of gas to be monitored in preset test environment and the test environment
The environmental testing data of the border factor.
Optionally, first computing module can be used for：By the content detection data and at least one ring
Border detection data is normalized.
First computing module can be used for：By the content detection data after the normalized and it is described at least
One environment measuring data input gas content computation model trained in advance.
By the way that the normalized of the content detection data and at least one environment measuring data can be made
Two kinds of data belong to same type of data, and there is no the situations of data class or Type-Inconsistencies, more convenient when calculating.
In the embodiment of the present invention, in the content detection data of gas to be monitored in monitoring current environment, also directed to environment
In other envirment factors be detected, obtain the environment measuring data of at least one envirment factor, the content detection data with
And the environment measuring data be input to the gas content computation model after can obtain gas to be monitored in current environment
In content monitoring data.The content monitoring data can improve the accuracy of monitoring due to consideration that various environmental factors.
As one embodiment, at least one envirment factor can include：Environment temperature, ambient humidity and interference
Gas；
First determining module includes：
Data determination unit, for determine the content detection data for the gas to be monitored for being acquired in current environment acquisition with
And the interference of the temperature detection data of the environment temperature, the Humidity Detection data of the ambient humidity and the interference gas
Content data.
The gas determination unit can include：
Determine in current environment using first gas sensor acquisition obtain gas to be monitored content detection data,
The Humidity Detection data and second gas that the temperature detection data of temperature sensor acquisition acquisition, humidity sensor acquisition obtain
The interference content data that sensor acquisition obtains.
The first gas sensor, temperature sensor, humidity sensor and second gas sensor may be used
Above-mentioned MEMS sensor can be specifically integrated into an array by MEMS sensor, and form MEMS sensor array is made with facilitating
With.
In the embodiment of the present invention, by different types of envirment factors such as temperature, humidity, the interference gas in environment
Under effect, the content for obtaining corresponding gas to be monitored can be calculated, considering the influence of the multiple environment factor can accurately define
The influence of different factor pairs gas to be monitored, obtains more accurate data influence result in environment.
As another embodiment, as shown in fig. 7, described device specifically can be by the way that with lower module, training obtains gas in advance
Body content calculation model：
Second determining module 701, for determining the corresponding at least one set of training data of at least one test environment；Wherein,
Each group of training data is included in the content preset data of the preset gas to be monitored in its corresponding test environment, described
The environment preset data of at least one envirment factor and the survey obtained in the test environment by sensor array acquisition
Try data；
Model construction module 702, for building gas content computation model；
Parameter training module 703, for being based at least one set of training data, training obtains the gas content and calculates
The model parameter of model.
At least one test environment refers to, pre-set the content of each gas and the numerical value of envirment factor with
For the test environment of test.The corresponding content preset data of the test environment is the content of preset gas to be monitored
Data, environment preset data are the data of other preset envirment factors.At least one test environment is according to
Content preset data and environment preset data and build what is finished.At least one envirment factor can be in above-described embodiment
Described environment temperature, ambient humidity and interference gas etc., when setting test environment, can set respectively environment temperature,
The content of ambient humidity, interference gas and gas to be monitored, to build corresponding test environment.
By MEMS sensor array when detection devices are placed in each test environment, the MEMS sensor array
Wait detection devices that can test the content for monitoring gas and at least one environment in the environment in each test environment
The numerical value of the factor namely the test data that the test environment can be determined by sensor arrays such as MEMS, obtain various gas
The test data of body and envirment factor.The test data can include gas content test data and environmental testing number
According to.
The gas content computation model refers to the mathematical model that can be used for calculating gas content.As a kind of possible
Realization method, the gas content computation model can refer to LSSVM (Least Squares Support Vector
Machine, least square method supporting vector machine) model, LSSVM models can include multiple parameters.The structure gas content meter
The multiple parameter values that model specifically refers to determine the LSSVM models are calculated, determine to make by the multiple parameter values for determining to use
The calculation formula of LSSVM at this point, determining the multiple parameter values used as unknown number, needs to utilize at least one set of training number
According to training obtains the model parameter of the gas content computation model.
As another embodiment, the parameter training module can include：
Reference parameter unit, for determining at least one set of reference parameter of the gas content computation model.
Content prediction unit, for will the content preset data and the environment preset data input described at least one
In the inversion model of the corresponding gas content computation model of group reference parameter, it is corresponding at least one pre- to obtain institute's test environment
Measured data.
Wherein, at least one preset data can refer to is obtained by the inversion model of gas content computation model by calculating
The content prediction data of gas to be monitored and the environmental forecasting number of at least one envirment factor in the test environment obtained
According to.
First comparing unit, at least one prediction data to be compared respectively with the test data.
The test data can include the content measuring data of gas to be monitored obtained by sensor array acquisition
And the test data of at least one envirment factor.It is described by least one prediction data respectively with the test data into
Row can relatively include：
The content prediction data of gas to be monitored with content measuring data are compared, obtain gas comparison result；
The environmental forecasting data of each envirment factor with environmental testing data are compared respectively, obtain environment
Comparison result.
More targetedly comparison result can be obtained by being respectively compared, for respectively from gas and envirment factor
Angle weighs the difference of test data and prediction data.
It is described that at least one prediction data is compared with the test data can also include respectively：
By the content prediction data in prediction data and at least one environmental forecasting data, with the content in test data
Test data and at least one environmental testing data, are carried out at the same time and compare, and obtain whole comparison result.
Pass through the whole whole comparison result that can relatively obtain prediction data and test data so that comparison result has more
Whole difference.
First selecting unit, for based on comparative result, optimum prediction number being selected from least one prediction data
According to.
First judging unit, for judging whether the optimum prediction data meet required precision.
First determination unit, for if so, determining the corresponding one group of object reference parameter of the optimum prediction data；It will
Model parameter of the object reference parameter as the gas content computation model.
Second determination unit, for if not, based on the object reference parameter, obtaining at least one set of genetic parameter；It will
At least one set genetic parameter returns to the determining gas content and calculates as at least one set of reference parameter
The step of at least one set of reference parameter of model, continues to execute.
Optionally, when the gas content computation model refers specifically to the LSSVM models, the model parameter of the LSSVM
Penalty factor and kernel function can specifically be referred to, the penalty factor and kernel function can be used as described in the embodiment of the present application
Reference parameter.
Determining at least one set of reference parameter can refer to, and according to needs are used, the parameter is carried out Initialize installation.
When the gas content computation model refers specifically to the LSSVM models, set at least one set of reference parameter that can specifically refer to described
The initialization data of penalty factor and kernel function, respectively described penalty factor and kernel function set an initialization data
When, you can to determine one group of initiation parameter.
The gas content computation model refers to what is obtained using sensor array acquisition, with the gas content to be monitored
The corresponding electric signal output value of the environment measuring data of detection data and at least one envirment factor, by the telecommunications
Number output valve can obtain the content monitoring data of the gas to be monitored and the environmental monitoring number of envirment factor by calculating
According to computation model.
Similarly, the environment of the content preset data of the gas to be monitored and at least one envirment factor is preset
Data when inputting the inversion model of the gas content computation model, can calculate acquisition and contain with what the sensor array acquired
The electric signal output for measuring test data and environmental testing data is worth corresponding content prediction data and environmental forecasting data.
It is described to judge whether the optimum prediction data meet required precision and refer to, by the optimum prediction data and survey
Data input precision calculation formula is tried, judges whether the precision difference of the two meets preset condition, is preset for example, mean square deviation is less than
Threshold value when meeting, that is, meets required precision.
By setting at least one set of reference parameter, using the reference parameter as training basis, based on the gas content
The inversion model of computation model carries out model training, screens legal optimum prediction data, expires in the optimum prediction data
During sufficient required precision, you can to determine corresponding object reference parameter, to determine corresponding object reference model as model parameter.
If the optimum prediction data are unsatisfactory for required precision, object reference parameter can be directed to as determining corresponding at least one
Group genetic parameter redefines corresponding model parameter.The comparison of prediction data and test data may be constructed the of model parameter
One layer of selection criteria, the second layer selection criteria for being selected as model parameter selection of required precision, bilayer selection may insure
Select corresponding more accurate model parameter.
In certain embodiments, the parameter training model further includes：
Number accumulated unit performs cumulative number for recording；
Second determination unit includes：
Judgment sub-unit, for judging currently to perform whether cumulative number is more than maximum execution number.
First determination subelement, for if so, determining the corresponding one group of object reference parameter of the optimum prediction data；
Using the object reference parameter as the model parameter of the gas content computation model.
Second determination subelement, for if not, based on the object reference parameter, obtaining at least one set of genetic parameter；
Using at least one set of genetic parameter as at least one set of reference parameter, the determining gas content meter is returned to
The step of at least one set of reference parameter for calculating model, continues to execute.
In above-described embodiment, when optimum prediction data are unsatisfactory for required precision, need to carry out model parameter hereditary excellent
Change and calculate.But genetic optimization calculating is an iterative calculation, limit perform number can to avoid occurring iterative process always,
Iterative calculation performs and can not obtain model parameter always.It can be with the continuity of guarantee procedure.
In certain embodiments, first comparing unit includes：
Fitness computation subunit, for being based on fitness function, respectively by each prediction data and the test number
It is calculated according to the fitness function is inputted, obtains at least one fitness value；
The first selecting unit includes：
Optimal selection subelement for being based at least one fitness value, determines at least one fitness value
In maximum adaptation angle value.
Target determination subelement, for determining that the corresponding prediction data of the maximum adaptation degree is the optimum prediction number
According to.
Based on fitness function, by least one sensor array prediction data and the Sensor array number
It is calculated according to the fitness function is inputted, to obtain at least one fitness value, by pair of the fitness value maximum of acquisition
The prediction data answered, as the optimum prediction data.
In above-described embodiment, at least one prediction data and test data are evaluated by using fitness function
Difference can improve the computational accuracy of difference by the use of fitness function, to select optimal model parameter, obtain compared with
The monitored results of good gas content computation model, the then final more accurate gas to be monitored of acquisition.
In certain embodiments, second determination unit includes：
Third determination subelement, for using the object reference parameter as genetic operator, institute to be obtained using genetic algorithm
State at least one set of genetic parameter.
It is described using the object reference parameter as genetic operator, using in genetic algorithm selection operation, crossover operation,
After mutation operation, more newly-generated a new generation population namely at least one set of genetic parameter, the genetic parameter may act as newly
At least one reference parameter of a generation.
The genetic algorithm can specifically refer to (Genetic Algorithm, GA algorithm), refer specifically to model Darwin biology
The computation model of the biological evolution process of evolutional natural selection and genetic mechanisms.The computation model combination above-described embodiment
Described in LSSVM algorithms when, each step of above-described embodiment still can be utilized to carry out the model parameter of gas to be monitored
It calculates, obtains the gas content computation model that GA-LSSVM is formed.
In above-mentioned realization method, by reference to the setting of parameter, determining for the gas content computation model can be realized,
By the inverse mould that preset gas content preset data and environment preset data are input to corresponding gas content computation model
In type, calculated by the inversion model of the gas content computation model after obtaining prediction data, by the prediction data with surveying
Examination data compare, and to determine optimal reference parameter, and then obtain the object reference parameter.
In the embodiment of the present invention, pass through the corresponding at least one set of test data of preset at least one test environment, training
The gas content computation model, obtains the model parameter of the gas content computation model, and model parameter is based at least one
The initial model parameter training of group obtains, and by repeatedly training, can obtain accurate model parameter so that the gas content
Computation model more levels off to accurate content calculation precision, accordingly, can improve containing for gas to be monitored in the embodiment of the present invention
The result of calculation of monitoring data is measured, to obtain accurate environmental monitoring result.
In a typical configuration, computing device includes one or more processors (CPU), input/output interface, net
Network interface and memory.
Memory may include computer-readable medium in volatile memory, random access memory (RAM) and/or
The forms such as Nonvolatile memory, such as read-only memory (ROM) or flash memory (flash RAM).Memory is computer-readable medium
Example.
Computer-readable medium includes permanent and non-permanent, removable and non-removable media can be by any method
Or technology come realize information store.Information can be computer-readable instruction, data structure, the module of program or other data.
The example of the storage medium of computer includes, but are not limited to phase transition internal memory (PRAM), static RAM (SRAM), moves
State random access memory (DRAM), other kinds of random access memory (RAM), read-only memory (ROM), electric erasable
Programmable read only memory (EEPROM), fast flash memory bank or other memory techniques, CD-ROM read-only memory (CD-ROM),
Digital versatile disc (DVD) or other optical storages, magnetic tape cassette, the storage of tape magnetic rigid disk or other magnetic storage apparatus
Or any other non-transmission medium, available for storing the information that can be accessed by a computing device.It defines, calculates according to herein
Machine readable medium does not include the data-signal and carrier wave of non-temporary computer readable media (transitory media), such as modulation.
Some vocabulary has such as been used to censure specific components in specification and claim.Those skilled in the art should
It is understood that hardware manufacturer may call same component with different nouns.This specification and claims are not with name
The difference of title is used as the mode for distinguishing component, but is used as the criterion of differentiation with the difference of component functionally.Such as logical
The "comprising" of piece specification and claim mentioned in is an open language, therefore should be construed to " include but do not limit
In "." substantially " refer in receivable error range, those skilled in the art can be described within a certain error range solution
Technical problem basically reaches the technique effect.In addition, " coupling " word is herein comprising any direct and indirect electric property coupling
Means.Therefore, if it is described herein that a first device is coupled to a second device, then representing the first device can directly electrical coupling
It is connected to the second device or the indirectly electrically coupled through other devices or coupling means second device.Specification
Subsequent descriptions be implement the application better embodiment, so it is described description be for the purpose of the rule for illustrating the application,
It is not limited to scope of the present application.The protection domain of the application is when subject to appended claims institute defender.
It should also be noted that, term " comprising ", "comprising" or its any other variant are intended to nonexcludability
Comprising so that commodity or system including a series of elements not only include those elements, but also including without clear and definite
It the other element listed or further includes as this commodity or the intrinsic element of system.In the feelings not limited more
Under condition, the element that is limited by sentence "including a ...", it is not excluded that in the commodity including the element or system also
There are other identical elements.
Several preferred embodiments of the application have shown and described in above description, but as previously described, it should be understood that the application
Be not limited to form disclosed herein, be not to be taken as the exclusion to other embodiment, and available for various other combinations,
Modification and environment, and the above teachings or related fields of technology or knowledge can be passed through in application contemplated scope described herein
It is modified.And changes and modifications made by those skilled in the art do not depart from spirit and scope, then it all should be in this Shen
It please be in the protection domain of appended claims.
Claims (14)
1. a kind of gas detection method, which is characterized in that including：
Determine the content detection data and at least of gas to be monitored obtained in current environment by sensor array acquisition
The environment measuring data of one envirment factor；
By the content detection data and at least one environment measuring data input gas content that training obtains in advance
Computation model calculates the content monitoring data for obtaining the gas to be monitored in current environment；
Wherein, the content monitoring data is used to assess whether the gas to be monitored pollutes current environment；The gas content
Content preset data, at least one envirment factor of the computation model based on the gas to be monitored preset in test environment
Environment preset data and in the test environment by sensor array acquisition obtain test data in advance training obtain
.
2. according to the method described in claim 1, it is characterized in that, the gas content computation model specifically as follows
Training obtains in advance：
Determine the corresponding at least one set of training data of at least one test environment；Wherein, it is right to be included in its for each group of training data
The content preset data of the preset gas to be monitored, the environment of at least one envirment factor are pre- in the test environment answered
If data and the test data obtained in the test environment by sensor array acquisition；
Build gas content computation model；
Based at least one set of training data, training obtains the model parameter of the gas content computation model.
3. according to the method described in claim 2, it is characterized in that, described based at least one set of training data, training obtains
The model parameter for obtaining the gas content computation model includes：
Determine at least one set of reference parameter of the gas content computation model；
The content preset data and environment preset data input at least one set of reference parameter is corresponding
In the inversion model of gas content computation model, the corresponding at least one prediction data of institute's test environment is obtained；
At least one prediction data is compared respectively with the test data；
Based on comparative result, optimum prediction data are selected from least one prediction data；
Judge whether the optimum prediction data meet required precision；
If so, determine the corresponding one group of object reference parameter of the optimum prediction data；Using the object reference parameter as
The model parameter of the gas content computation model；
If not, based on the object reference parameter, at least one set of genetic parameter is obtained；At least one set of genetic parameter is made
For at least one set of reference parameter, at least one set of reference parameter for determining the gas content computation model is returned to
The step of continue to execute.
4. according to the method described in claim 3, it is characterized in that, it is described using at least one set of genetic parameter as described in extremely
Few one group of reference parameter, return to it is described determine the gas content computation model at least one set of reference parameter the step of after
After continuous execution, the method further includes：
Record performs cumulative number；
It is described to be based on the object reference parameter, obtain at least one set of genetic parameter；Using at least one set of genetic parameter as
At least one set reference parameter, returns at least one set of reference parameter for determining the gas content computation model
Step continue to execute including：
Judge whether current execution cumulative number is more than maximum and performs number；
If so, determine the corresponding one group of object reference parameter of the optimum prediction data；Using the object reference parameter as
The model parameter of the gas content computation model；
If not, based on the object reference parameter, at least one set of genetic parameter is obtained；At least one set of genetic parameter is made
For at least one set of reference parameter, at least one set of reference parameter for determining the gas content computation model is returned to
The step of continue to execute.
5. according to the method described in claim 3, it is characterized in that, it is described by least one prediction data respectively with it is described
Test data be compared including：
Based on fitness function, each prediction data and the test data are inputted into the fitness function respectively and carried out
It calculates, obtains at least one fitness value；
It is described based on comparative result, from least one prediction data select optimum prediction data include：
Based at least one fitness value, the maximum adaptation angle value at least one fitness value is determined；
It is the optimum prediction data to determine the corresponding prediction data of the maximum adaptation degree.
It is 6. according to the method described in claim 3, it is characterized in that, described based on the object reference parameter, acquisition at least one
Group genetic parameter includes：
Using the object reference parameter as genetic operator, at least one set of genetic parameter is obtained using genetic algorithm.
7. according to the method described in claim 1, it is characterized in that, at least one envirment factor includes：Environment temperature, ring
Border humidity and interference gas；
The content detection data of the gas to be monitored of acquisition and at least one envirment factor are acquired in the determining current environment
Environment measuring data include：
It determines to acquire the content detection data of the gas to be monitored of acquisition and the temperature of the environment temperature in current environment
The interference content data of detection data, the Humidity Detection data of the ambient humidity and the interference gas.
8. a kind of gas-detecting device, which is characterized in that including：
First determining module, for determining content detection data for the gas to be monitored for being acquired in current environment acquisition and extremely
The environment measuring data of a few envirment factor；
First computing module, for the content detection data and at least one environment measuring data input to be instructed in advance
Practice the gas content computation model obtained, calculate the content monitoring data for obtaining the gas to be monitored in current environment；
Wherein, the content monitoring data is used to assess whether the gas to be monitored pollutes current environment；The gas content
Content preset data, at least one envirment factor of the computation model based on the gas to be monitored preset in test environment
Environment preset data and in the test environment by sensor array acquisition obtain test data in advance training obtain
.
9. device according to claim 8, which is characterized in that described device is especially by training obtains in advance with lower module
Gas content computation model：
Second determining module, for determining the corresponding at least one set of training data of at least one test environment；Wherein, each group of instruction
Practice data and be included in the content preset data of the preset gas to be monitored in its corresponding test environment, described at least one
The environment preset data of envirment factor and the test data obtained in the test environment by sensor array acquisition；
Model construction module, for building gas content computation model；
Parameter training module, for being based at least one set of training data, training obtains the gas content computation model
Model parameter.
10. device according to claim 9, which is characterized in that the parameter training module includes：
Reference parameter unit, for determining at least one set of reference parameter of the gas content computation model；
Content prediction unit, for the content preset data and the environment preset data input at least one set to be joined
In the inversion model for examining the corresponding gas content computation model of parameter, the corresponding at least one prediction number of institute's test environment is obtained
According to；
First comparing unit, at least one prediction data to be compared respectively with the test data；
First selecting unit, for based on comparative result, optimum prediction data being selected from least one prediction data；
First judging unit, for judging whether the optimum prediction data meet required precision；
First determination unit, for if so, determining the corresponding one group of object reference parameter of the optimum prediction data；By described in
Model parameter of the object reference parameter as the gas content computation model；
Second determination unit, for if not, based on the object reference parameter, obtaining at least one set of genetic parameter；By described in
At least one set of genetic parameter returns to the determining gas content computation model as at least one set of reference parameter
At least one set of reference parameter the step of continue to execute.
11. device according to claim 10, which is characterized in that the parameter training model further includes：
Number accumulated unit performs cumulative number for recording；
Second determination unit includes：
Judgment sub-unit, for judging currently to perform whether cumulative number is more than maximum execution number；
First determination subelement, for if so, determining the corresponding one group of object reference parameter of the optimum prediction data；By institute
State model parameter of the object reference parameter as the gas content computation model；
Second determination subelement, for if not, based on the object reference parameter, obtaining at least one set of genetic parameter；By institute
At least one set of genetic parameter is stated as at least one set of reference parameter, the determining gas content is returned to and calculates mould
The step of at least one set of reference parameter of type, continues to execute.
12. device according to claim 9, which is characterized in that first comparing unit includes：
Fitness computation subunit, it is respectively that each prediction data and the test data is defeated for being based on fitness function
Enter the fitness function to be calculated, obtain at least one fitness value；
The first selecting unit includes：
Optimal selection subelement for being based at least one fitness value, is determined at least one fitness value
Maximum adaptation angle value；
Target determination subelement, for determining that the corresponding prediction data of the maximum adaptation degree is the optimum prediction data.
13. device according to claim 9, which is characterized in that second determination unit includes：
Third determination subelement, for using the object reference parameter as genetic operator, using genetic algorithm obtain described in extremely
Few one group of genetic parameter.
14. device according to claim 8, which is characterized in that at least one envirment factor includes：Environment temperature,
Ambient humidity and interference gas；
First determining module includes：
Data determination unit, for determining the content detection data for the gas to be monitored for being acquired in current environment acquisition and institute
State the interference content of the temperature detection data of environment temperature, the Humidity Detection data of the ambient humidity and the interference gas
Data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711436911.9A CN108152458A (en) | 2017-12-26 | 2017-12-26 | Gas detection method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711436911.9A CN108152458A (en) | 2017-12-26 | 2017-12-26 | Gas detection method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108152458A true CN108152458A (en) | 2018-06-12 |
Family
ID=62461993
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711436911.9A Pending CN108152458A (en) | 2017-12-26 | 2017-12-26 | Gas detection method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108152458A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109858144A (en) * | 2019-01-29 | 2019-06-07 | 江苏啸峰环保科技股份有限公司 | A kind of the discharge capacity account model building method and fugitive dust discharge capacity account method in fugitive dust source |
| CN110617916A (en) * | 2019-09-29 | 2019-12-27 | 歌尔股份有限公司 | Calibration method and device of air pressure sensor |
| CN113281472A (en) * | 2021-07-26 | 2021-08-20 | 北京英视睿达科技有限公司 | Ambient air quality sensor, monitoring device and method |
| CN113791128A (en) * | 2021-11-16 | 2021-12-14 | 南京尚中过滤与分析设备有限公司 | Safety alarm system and method for gas concentration in converter gas |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105444923A (en) * | 2015-11-18 | 2016-03-30 | 浙江工业大学 | Mechanical temperature instrument error prediction method based on genetic-algorithm optimized least square support vector machine |
| CN106405007A (en) * | 2016-08-30 | 2017-02-15 | 河北先河环保科技股份有限公司 | Novel calibration method for gas sensors and particulate matter sensors |
| CN106645593A (en) * | 2017-01-24 | 2017-05-10 | 浙江农林大学 | Hazardous chemical poison reagent leakage detection method |
| CN107085018A (en) * | 2017-04-27 | 2017-08-22 | 浙江清华长三角研究院 | A kind of electric nasus system detected for VOC and stench class gas |
| CN105891422B (en) * | 2016-04-08 | 2017-08-25 | 重庆大学 | The electronic nose Gas Distinguishing Method that the limit learns drift compensation is migrated based on source domain |
-
2017
- 2017-12-26 CN CN201711436911.9A patent/CN108152458A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105444923A (en) * | 2015-11-18 | 2016-03-30 | 浙江工业大学 | Mechanical temperature instrument error prediction method based on genetic-algorithm optimized least square support vector machine |
| CN105891422B (en) * | 2016-04-08 | 2017-08-25 | 重庆大学 | The electronic nose Gas Distinguishing Method that the limit learns drift compensation is migrated based on source domain |
| CN106405007A (en) * | 2016-08-30 | 2017-02-15 | 河北先河环保科技股份有限公司 | Novel calibration method for gas sensors and particulate matter sensors |
| CN106645593A (en) * | 2017-01-24 | 2017-05-10 | 浙江农林大学 | Hazardous chemical poison reagent leakage detection method |
| CN107085018A (en) * | 2017-04-27 | 2017-08-22 | 浙江清华长三角研究院 | A kind of electric nasus system detected for VOC and stench class gas |
Non-Patent Citations (5)
| Title |
|---|
| ZHOU HUAPING ET AL.: "Application of Support Vector Machine Model in Mine Gas Safety Level Prediction", 《INDONESIAN JOURNAL OF ELECTRICAL ENGINEERING》 * |
| 宋凯: "金属氧化物半导体气体传感器气体检测关键问题研究", 《中国博士学位论文全文数据库 信息科技辑》 * |
| 董海鹰 编著: "《智能控制理论及应用》", 30 September 2016, 中国铁道出版社 * |
| 靳蕃 等主编: "《神经网络理论与应用研究‘96》", 31 October 1996, 西南交通大学出版社 * |
| 黄银蓉: "污水处理曝气池溶解氧智能优化控制系统研究", 《中国优秀硕士学位论文全文数据库 工程科技Ⅰ辑》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109858144A (en) * | 2019-01-29 | 2019-06-07 | 江苏啸峰环保科技股份有限公司 | A kind of the discharge capacity account model building method and fugitive dust discharge capacity account method in fugitive dust source |
| CN110617916A (en) * | 2019-09-29 | 2019-12-27 | 歌尔股份有限公司 | Calibration method and device of air pressure sensor |
| CN110617916B (en) * | 2019-09-29 | 2021-08-31 | 潍坊歌尔微电子有限公司 | Calibration method and device of air pressure sensor |
| CN113281472A (en) * | 2021-07-26 | 2021-08-20 | 北京英视睿达科技有限公司 | Ambient air quality sensor, monitoring device and method |
| CN113791128A (en) * | 2021-11-16 | 2021-12-14 | 南京尚中过滤与分析设备有限公司 | Safety alarm system and method for gas concentration in converter gas |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108152458A (en) | Gas detection method and device | |
| US8650137B2 (en) | Method and apparatus for creating state estimation models in machine condition monitoring | |
| Nunzi et al. | Detection of anomalies in the behavior of atomic clocks | |
| CN106441632B (en) | Detection method of thermometer and detection device of thermometer | |
| CN111104736B (en) | Abnormal data detection method, device, medium and equipment based on time sequence | |
| CN114492629A (en) | Abnormality detection method, abnormality detection device, electronic apparatus, and storage medium | |
| CN114707632A (en) | Sensor network sensor fault positioning method, system, equipment and medium | |
| CN115824993B (en) | Method and device for determining water body chemical oxygen demand, computer equipment and medium | |
| CN104768175B (en) | A kind of Local Area Network coverage detection method and system | |
| CN111881800A (en) | Concrete filled steel tube void defect detection method based on AR model and self-service inspection | |
| CN116242899B (en) | Method for calibrating electrode method water quality detection result based on machine learning | |
| CN113190985B (en) | Analysis method and system for data errors in layered heterogeneous region | |
| CN111859744B (en) | Method, device and equipment for identifying node rigid domain based on monitoring data | |
| RU2747739C1 (en) | Verification method for sea water salinity measurements by automatic measuring instruments | |
| KR101511289B1 (en) | Method, apparatus and system for estimating degree of uniformity of susceptor | |
| CN108332722A (en) | A kind of sea water advanced detection method and device | |
| CN110705186B (en) | Real-time online instrument checksum diagnosis method through RBF particle swarm optimization algorithm | |
| Dillingham et al. | Characterising uncertainty in instrumental limits of detection when sensor response is non-linear | |
| CN112577671A (en) | Well lid monitoring method and system by using Kriging method | |
| Kouadri et al. | A statistical-based approach for fault detection in a three tank system | |
| CN104931653A (en) | Calibration method of pollution accumulation difficulty level of suspension insulator based on equivalent salt deposit density parameters | |
| CN114383834A (en) | Ocean engineering structure micro-damage judgment method | |
| CN112241343A (en) | Slow disk detection method and device, electronic equipment and readable storage medium | |
| CN103257206A (en) | Evaluation method of detection result accuracy of rapid organophosphorus and carbamate pesticide residue detector | |
| CN115308517B (en) | Aging detection method and system for components, storage medium and equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20200609 Address after: 261031 building 10, Geer phase II Industrial Park, No. 102, Ronghua Road, Ronghua community, Xincheng street, high tech Zone, Weifang City, Shandong Province Applicant after: Weifang goer Microelectronics Co.,Ltd. Address before: 261031 No. 268 Dongfang Road, Weifang hi tech Industrial Development Zone, Shandong, Weifang Applicant before: GOERTEK Inc. |
|
| TA01 | Transfer of patent application right | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180612 |
|
| RJ01 | Rejection of invention patent application after publication |