CN105388884A - Alarm system for detecting leakage fault of heat supply network based on identification algorithm driven by data and method - Google Patents
Alarm system for detecting leakage fault of heat supply network based on identification algorithm driven by data and method Download PDFInfo
- Publication number
- CN105388884A CN105388884A CN201510749968.9A CN201510749968A CN105388884A CN 105388884 A CN105388884 A CN 105388884A CN 201510749968 A CN201510749968 A CN 201510749968A CN 105388884 A CN105388884 A CN 105388884A
- Authority
- CN
- China
- Prior art keywords
- data
- pipe network
- network
- leakage failure
- real
- 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
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B23/00—Testing or monitoring of control systems or parts thereof
- G05B23/02—Electric testing or monitoring
- G05B23/0205—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults
- G05B23/0208—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterized by the configuration of the monitoring system
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Automation & Control Theory (AREA)
- Examining Or Testing Airtightness (AREA)
Abstract
The invention relates to an alarm system for detecting a leakage fault of a heat supply network based on an identification algorithm driven by data and a method. According to real-time monitoring data of pressures of an inlet and an outlet of a circulating water pump or a user side, data are converted to a certain extent according to an input requirement of an identification algorithm; a real-time operation state of a heat supply network in a normal operation mode or a leakage fault mode is identified; and identification and detection are carried out. In addition, the system comprises a basic heat-supply pipe network system, a data collection system, a communication system, a pipe network operation management system, and an alarm system; and the heat-supply pipe network system is connected with a circulating water pump and a pressure gauge or a wireless pressure sensor at a user side. The data collection system is connected with the communication system to transmit the collected data to the pipe network operation management system by using a GPRS as a medium. After data pretreatment and feature extraction, the system carries out identification; and then a command is sent to the alarm system based on an identification result. Therefore, automatic adjustment and improvement of an identification rule are realized; the accuracy of the identification algorithm is improved; and false alarm occurrence is reduced.
Description
Technical field
The invention belongs to the crossing domain of HV&AC engineering and infotech etc.Propose a kind of warning system and the method that only just can detect heating network leakage failure based on the import and export pressure monitoring data of water circulating pump or user side and recognizer.Be specially a kind of warning system and the method that detect heating network leakage failure based on the recognizer by data-driven.
Background technology
Along with expansion and the growth in operation age of central heating system scale, the continuous generation of China's big or middle city thermal pipe network leakage accident, wherein common with leakage failure.If when detection and location after pipe leakage and maintenance not in time, the serious wasting of resources and economic loss will be caused, also can cause the pollution of environment, or even human casualty accident.As an important component part of reliability of heating network, the leakage failure diagnosis carrying out heat supply network is the effective means ensureing heating network economy, safe operation and the automatic control of raising heating network, management level.
Fault diagnosis refers to the process whether certainty annuity breaks down, namely to the testing process of an abnomal condition.By the change of continuous monitoring system measurable variable, to export according to system or the characteristic of estimation residual error of state variable judges.In conjunction with the method for current heat distribution pipe network fault diagnosis both at home and abroad, two large classes can be divided into: based on hardware diagnostic method with based on diagnostic software code [1].Mainly rely on manual inspection based on hardware diagnostic method or utilize the principle of various physics and chemistry, the various physics and chemistry phenomenons occurred with fault by direct-detection are to realize fault diagnosis.Method for diagnosing faults based on software mainly relies on the change detecting the characterisitic parameter such as pressure, flow and temperature when heat distribution pipe network occurs to leak, and utilizes specific algorithm to carry out fault leak diagnostics and location.
Hardware based fault diagnosis mainly contains the methods such as manual inspection, calorifics diagnosis, acoustics diagnose, leakage flux diagnosis and distribution type fiber-optic diagnosis, wherein by based on the diagnostic techniques of acoustics and the combination of signal analysis method, improve the verification and measurement ratio [2] of heat supply network leakage failure.Although hardware based fault detection technique sensitivity is good, accurate positioning, cost higher, actual pipe network generation leakage situation cannot be predicted.So hardware based fault diagnosis technology is general not as the Main Means of heat distribution pipe network fault diagnosis.
Fault diagnosis (based on signal transacting, based on analytic model with based on artificial intelligence) method based on software becomes the main flow of accident of pipeline network just gradually.Diagnostic method based on signal transacting mainly contains negative pressure wave method and spot pressure analytic approach, and the negative pressure wave signal that negative pressure wave method utilizes pipe leakage place to occur that instantaneous pressure bust is formed carries out leakage and judges.Although this method is for leak detection and be positioned with very high degree of accuracy, the characteristic of suction wave determines that it is only suitable for fast and the leakage of burst, does not have the situation of obvious suction wave then invalid [3] very slowly, be often used to leaking hunting of oil pipeline for leakage rate.Whether the generation that spot pressure analytic approach utilizes the curve comparison when method of statistics extraction pipeline pressure history along the line and pipeline normal operating condition to judge to leak, because the method only needs the pressure signal of one or several check points, so the response time is shorter, calculated amount is little, but to the evaluation capacity of leakage rate poor [4].Diagnosis based on analytic model needs the mathematical model setting up heat supply network hydraulic analysis, the distribution of pipe flow field is solved under given boundary condition, actual measured value and model calculation value are contrasted locating leaks point, mainly contains based on state estimate with based on Parameter Estimation Method [5] [6].The former detection method not only requires that the precision type of surveying instrument is high, and is only applicable to the accident of pipeline network diagnosis of Small leak.Although the latter has better practicality and accuracy, algorithm is complicated, and the response time is slow, and when heat supply network working conditions change is little, very easily occurs not restrain situation.Method based on artificial intelligence mainly refers to the method based on neural network, and the method that neural net method mainly utilizes nerual network technique to combine with statistical technique or genetic algorithm or evidence theory etc. carries out fault diagnosis [7].Although the method based on artificial intelligence has higher verification and measurement ratio, lower wrong report and rate of failing to report, due to the limitation of training data, can only for some leakage situation, poor to the assessment of actual pipeline network leak fault.
In true heating network, pipe network is much far complicated than general simplified model, and the pressure of often just monitoring water circulating pump and the import and export of user's end and data on flows, there is no more signal or Data Collection.Thus, take a broad view of above heat supply network leakage fault diagnosing method, all have some limitations: the diagnostic method based on Data Analysis is difficult to set up hot net water force analysis model accurately, based on the diagnostic method of signal transacting then owing to cannot use without pressure wave data, these all can cause the wrong report of pipeline network leak accident or fail to report.Therefore, a kind of leakage failure detection system based on existing Monitoring Data being applicable to heating network of exploitation is needed badly.
[1] Zhao Kai. heat supply network Research on fault diagnosis method [D]. North China Electric Power University, 2012.
[2]SatoT,MitaA.Leakdetectionusingthepatternofsoundsignalsinwatersupplysystems[C]//The14thInternationalSymposiumon:SmartStructuresandMaterials&NondestructiveEvaluationandHealthMonitoringInternationalSocietyforOpticsandPhotonics,2007:65292K-65292K-9.
[3] Fu Daoming, Sun Jun, He Zhigang, etc. Summary of Pipeline Leakage Detection Technology [J] both at home and abroad. petroleum machinery, 2004,03 phase (3): 48-51.DOI:doi:10.3969/j.issn.1001-4578.2004.03.018.
[4]Gamboa-MedinaMM,ReisLFR,GuidoRC.FeatureExtractioninPressureSignalsforLeakDetectioninWaterNetworks[J].ProcediaEngineering,2014,70:688–697.
[5]DanetiM.Onusingphasedatainformationforpipelineleaklocation[C]//ElectricalandElectronicsEngineersinIsrael,2008.IEEEI2008.IEEE25thConventionofIEEE,2008:494-498.
[6] Qin Xuzhong, Jiang Yi. the on-line identification of region heat supply network resistance of pipe system coefficient and fault diagnosis [J]. Tsing-Hua University's journal (natural science edition), 2000,02 phase (2): 81-85.DOI:doi:10.3321/j.issn:1000-0054.2000.02.022.
[7] Duan Lanlan, Tian Qi, Duan Pengfei, etc. based on the heat-supply network failure diagnostic model [J] of genetic optimization BP neural network. Northcentral University's journal: natural science edition, 2014,03 phase.
Summary of the invention
The present invention is intended to the method by data mining, leaks the characteristic quantity operating mode, develop the pipeline network leak Fault Identification algorithm had compared with high-accuracy, and invent a kind of Leak Detection and warning system based on this algorithm from the pressure monitoring extracting data of heating network.According to the Real-time Monitoring Data in pipe network operation process, the normally inlet and outlet pressure of water circulating pump and user side and data on flows, detects whether there is leakage situation, can be directly used in the daily operation management work of the central heating network of different scales.
The present invention proposes the warning system that a kind of import and export pressure monitoring data based on water circulating pump or user side and recognizer just can detect pipe network leakage failure.
The present invention adopts following technical scheme:
A kind of recognition methods that whether there is the data-driven of leakage for detecting heating network; It is characterized in that the import and export pressure Real-time Monitoring Data according to water circulating pump or user side, data are carried out certain conversion according to the input requirements of recognizer, pipe network real-time running state when identifying normal operation or leakage failure occurs, carries out identifying, detecting.
Pipe network real-time running state step during leakage failure: search on the time-serial position of pressure data in chronological order; The characteristic and recognition rule of each point on curve or each section is mated; Determine real-time status; If its feature meets the leakage failure feature in recognition rule, then think that this section there occurs leakage failure; Otherwise, determine that this section is normal operating condition.
The process of establishing of leaking recognition rule is: first, the Real-time Monitoring Data of pipe network is collected; After data prediction, be used to that algorithm is described to curve and train, verify and test process; Finally export the recognition rule of leakage failure.
Curve describes algorithm and carries out training, to verify and test process is: be divided into training set and test set in the data that pre-service is good; Subsequently, training set is used to training and the checking of recognizer.
A kind of warning system detecting heating network leakage failure based on the recognizer by data-driven of the present invention; Comprise basic unit's heat distribution pipe network system, data acquisition system (DAS) and communication system, pipe network operation management system, warning system; Heat distribution pipe network system connects tensimeter or the wireless pressure sensor of water circulating pump and user side, the data of collection are that medium is sent to pipe network operation management system with GPRS by data acquisition system (DAS) connecting communication system, pipe network operation management system is through carrying out the pre-service of data and the extraction of feature, identify, finally given an order to warning system by recognition result.
Be described as follows:
Heating network, it can be the pipe network (secondary network) of a heat exchange station scope, also can be the pipe network (first-level pipeline network) within the scope of a thermal substation, wherein each heat exchange station can be regarded as user side: a complete confession backwater circulation network can be regarded as this heating network described in invention.As the arbitrary pipeline section in heating network or user side hot water leak, comprise user carry out at heat supply end equipment place people for the heat supply network leakage failure such as to discharge water occur time, by the extraction of pressure monitoring data and the analysis of recognizer, the digital signal of recognition result is converted to the signal such as sound, light, send to warning system, complete the detection of heat supply network leakage failure.Wherein, recognizer is inspection center's layer, is the core of this system.
The extraction of described pressure monitoring data, the water circulating pump of acquiring heat supply pipe network or the confession (entering) of user's end, return the data such as tensimeter on (going out) water pipe or pressure transducer, by the mode of wired or wireless (internet) to the memory device of administrative center as mobile phone or computer transmission pressure data.Then according to pressure characteristic during image data extraction leakage failure, mainly leak to the import and export pressure of water circulating pump or user side produce be obviously different from normal operation time feature, during as leaked generation, variable quantity and the rate of change of the import and export pressure curve of water circulating pump and user side have significant change; Extracting pressure characteristic can adopt information gain in rote learning (data mining), attribute set to select or the method such as principal component analysis (PCA).
Described recognizer, have employed a kind of rote learning (data mining) sorting algorithm having supervision, can by the learning and training of a small amount of sample size, water circulating pump during extraction heating network generation leakage failure or the import and export pressure characteristic of user side, form leakage failure recognition rule, and then utilize recognition rule to carry out Real-Time Monitoring, and recognition result is presented to pipe network operation supvr on human-computer interaction interface.Wherein sorting algorithm, for curve describes algorithm, by describing the shape facility of curve, excavating the behavior pattern hidden in curve, forming the template (recognition rule) of AD HOC; And then, this template and real-time curve are contrasted, if segment of curve and this template match, then thinks and occurred this pattern in this segment of curve.This algorithm can run, as single-chip microcomputer, computer, server etc. on any equipment with data-handling capacity.Namely a small amount of sample size is the training set of sorting algorithm, gathers, or collect in the pipe network of actual motion by laboratory small size testing table; Leak sample (1 leakage failure can be used as 1 and leaks sample) as long as exist, can training set be formed; The leakage sample that after having trained, still sustainable collection is new, forms new training set, thus carries out calibrating to algorithm and improve.Recognition rule is the template curve or literal expression that are formed by the pressure characteristic during leakage failure extracted, for carrying out contrasting and mating with real-time pressure data curve, detecting and whether there is leakage failure generation.
Described recognition result, can adopt various ways to carry out stating and export.Possible form is by the real-time pressure data of water circulating pump or user and a pipe network operation state recognition result, and the man-machine interface of administrative center carries out visual presentation; Also can be directly carry out alarm in a voice form, once leakage failure be detected, will immediately at administrative center's horn blew; Also can report to the police by the mode of phone or note.
The present invention is different from the detection method of conventional alarm system:
1) do not need additionally to arrange numerous pressure or flow sensor, the present invention only utilizes the input quantity of data as algorithm of design of pipe networks and intrinsic tensimeter or Sensor monitoring when running;
2) do not need to set up heat supply network hydraulic analysis model that is complicated or that simplify, avoid the leakage predicated error that the gap between Hydraulic Analysis of Water Pipe Networks model and realistic model is brought, improve the accuracy of warning system;
3) the present invention only in pipe network debugging or operation phase, or in the undersized test run in laboratory, need collect the leakage failure data of about 10 times, carries out training and verifies, can set up the recognition rule that an accuracy is higher to algorithm;
4) in the present invention training sample can with the leakage failure in pipe network increase and continuous updating, automatic adjustment and the improvement of recognition rule can be realized, improve constantly the accuracy rate of recognizer, reduce wrong report and fail to report.
Accompanying drawing explanation
Fig. 1 pipeline network leak fault recognition method process flow diagram;
Fig. 2 detects the process flow diagram of pipe network leakage failure;
The process of establishing block diagram of Fig. 3 pipeline network leak Fault Identification rule;
The training of Fig. 4 recognizer, checking and testing process block diagram;
Fig. 5 curve describes the morpheme schematic diagram for data conversion in algorithm;
Fig. 6 curve describes a kind of expression-form of the recognition rule that algorithm is set up;
Fig. 7 pipeline network leak Fault Identification Sample Rules (extract from the import and export pressure data of water circulating pump and leak feature);
Fig. 8 pipeline network leak recognition result shows example (being directly presented on the time-serial position of import and export pressure of water circulating pump);
Fig. 9 is based on the Leak Detection of recognizer and warning system workflow diagram;
Figure 10 detects pipe network leakage failure warning system example.
Embodiment
Below in conjunction with drawings and Examples, the present invention is described in detail.
Equipment connection illustrates: comprise basic unit's heat distribution pipe network system, data acquisition system (DAS) and communication system, pipe network operation management system, warning system, heat distribution pipe network system connects tensimeter or the wireless pressure sensor of water circulating pump and user side, the data of collection are that medium is sent to pipe network operation management system with wireless network by data acquisition system (DAS) connecting communication system, pipe network operation management system is through carrying out the pre-service of data, the extraction of feature and the realization of recognizer, identify, signal to be changed by the communication converter that is connected to computer end and is transferred to alarm controller by recognition result, finally outputed signal by alarm controller, make sound and light alarm, telephone call or SMS alarm give the alarm, complete the realization of system of the present invention.
Methods of operating etc. are described as follows with reference to the accompanying drawings:
Be illustrated in figure 1 the recognition methods process flow diagram that was normally run or occurred the pipe network real-time running state of leakage failure.By the tensimeter be arranged on water circulating pump or user side import and export pipeline section or pressure transducer real-time data collection 101, the data 101 gathered are by being wirelessly transmitted to computer client, and utilize client data process software such as excel etc. to carry out data conversion 102 according to the input requirements of recognizer, data conversion is that data are carried out certain conversion according to the input requirements of recognizer, as standardization, if the pressure data used in this enforcement is to the conversion etc. of Curve Symbol chain, for use in the process of recognizer.Detect pipe network according to recognition methods subsequently normally run or leakage failure 103 occurs.
Be illustrated in figure 2 the workflow diagram detecting pipe network leakage failure 103, comprise 3 steps: on the time-serial position of pressure data, search for 201 in chronological order; The characteristic and recognition rule of each point on curve or each section is carried out coupling 202; Determine pipe network real-time running state 203.Searching in the process of 201 on the time-serial position of pressure data in chronological order, can carry out point by point search to pretreated data 102 or search for by segment of curve, the algorithm in the present embodiment have employed the way of search by segment of curve.After this, the characteristic and recognition rule of the segment of curve searched is mated, if its feature meets the leakage failure feature in recognition rule, then think that this section there occurs leakage failure; Otherwise, determine that this section is normal operating condition.
Be illustrated in figure 3 the process of establishing of the pipeline network leak recognition rule determining to use in pipeline network leak fault 103.First, the Real-time Monitoring Data 301 of pipe network is collected, and comprises the import and export pressure of water circulating pump and user side; After data prediction 302, be used to that algorithm is described to curve and train, verify and test process 303; Finally export the recognition rule 304 of leakage failure.Wherein, the content of data prediction 302 is similar to data conversion process 102, but an odd word, be carry out class mark to pressure data, data point during as normally run can be labeled as 0, and data point when running under leakage failure can be labeled as 1.
Be illustrated in figure 4 the training of 303 recognizers, checking and testing process.The data that pre-service is good in process of data preprocessing 302 are divided into training set 401 and test set 402.Subsequently, training set 401 is used to training and the checking 403 of recognizer.The training of algorithm and proof procedure 403, be actually the pressure characteristic under extraction leakage failure and set up the process of recognition rule.In the present embodiment, k rolls over cross validation method and is used to training and the checking work of algorithm, and object is to reduce the possibility of over-fitting (over ?fitting).Through training, candidate's recognition rule 405 can be obtained, utilize test set 402 to test this recognition rule, and assessment 406 is carried out to test result.If recognition accuracy can accept, as reached more than 90%, then can be used as recognition rule and export, otherwise, training process 403 need be got back to.
Be divided into normal operation and leakage failure operation two states because pipe network operation state is actually, therefore identified leakage fault is actually and these two kinds of running statuses is distinguished, i.e. an assorting process.In chronological order on the time curve of pressure data in search procedure 201, search is by pointwise or mode piecemeal, is actually to be determined by sorting algorithm.Common statistics class algorithm, as C4.5 decision tree, artificial neural network etc., is point by point search; And curve describes algorithm is search for piecemeal.The implementation case have employed the curve being more suitable for processing time sequence curve and describes algorithm.
If Fig. 5 is the morpheme schematic diagram 500 that curve describes for data conversion in algorithm, comprise basic morpheme 501, expansion morpheme 502, do not comprise expansion morpheme 503, and wildcard morpheme 504.Wherein, basic morpheme 501 comprises 3 kinds, ascent stage or point 0 (x (i)-x (i-1) >0), descending branch/1 (x (i)-x (i-1) <0) and smooth section/2 (x (i)-x (i-1)=0) respectively; Expansion morpheme comprises 2 kinds, peak point 3 and valley point 4; In the present embodiment, wildcard morpheme 504 is commonly used to define those segment of curve not needing to understand its specific features.
Be illustrated in figure 6 a kind of expression-form that curve describes the recognition rule that algorithm is set up.Curve describes algorithm first by pressure data time-serial position 601, according to the morpheme 500 of definition, is converted to symbolic link 602, and is described by the numerical characteristics of proper vector group 603 to this segment of curve.
In proper vector group 603, arbitrary proper vector is defined as F=(M, f
1, f
2..., f
n).Wherein, M is the morpheme value of each symbol (A-E as in pressure data time-serial position 601), f
1, f
2..., f
nbe used to the characteristic quantity of the numerical characteristics describing this symbol.Different symbols can have different characteristic quantities, chooses suitable characteristic quantity and can reduce the operand of algorithm and improve recognition accuracy.Characteristic quantity choose and threshold value is determined to be operated in the training of algorithm, proof procedure 303 and is completed.The characteristic quantity choosing method of information gain is have employed in the present embodiment.
Be illustrated in figure 7 pipeline network leak Fault Identification rule 304 examples.When leakage failure occurs, the import and export pressure curve of water circulating pump can produce the trend that significantly declines to a great extent simultaneously, can recover gradually subsequently because of moisturizing cause.So this section of curve is crawled is the template curve 701 of leakage mode 702, and sets up proper vector group 703 and 704 by training and test process.Wherein, the characteristic quantity chosen is the variable quantity d that water circulating pump enters point-to-point transmission on (going out) mouth pressure curve
aBand rate of change dr
aB.Template curve 701 and proper vector group 703 and 704 together constitute the recognition rule of leakage failure pattern, this rule also usable text is described, as shown in recognition rule establishment step 705 in text box: if meet the condition that recognition rule sets up 705, then think and there occurs leakage failure, otherwise think that pipe network is normal operation.
Be illustrated in figure 8 pipeline network leak recognition result and show example, recognition result 802 has directly been illustrated on the time-serial position 801 of the import and export pressure of water circulating pump.Like this can Timeliness coverage leakage failure, and can be observed the impact of leakage failure on the import and export pressure of water circulating pump.
As Fig. 9 illustrates Leak Detection based on recognizer and warning system workflow diagram, comprise basic unit's heat distribution pipe network system, data acquisition system (DAS) and communication system, pipe network operation management system and warning system.First by being arranged in the pressure data of the wireless pressure sensor transmission measurement of heat network system water circulating pump and user's import and export to data acquisition unit.If pipe network system more complicated, distributed data acquisition unit can transfer data to wireless data acquisition terminal and focus on.Then utilize means of communication such as wireless network the data of collection to be transferred to the realization that computer client or server carry out the pre-service of data, the extraction of feature and algorithm, concrete operation method is wherein described in detail above, no longer explains here.Finally being given an order arrival warning system by the recognition result of algorithm, can be detection and the alarm that sound and light alarm, telephone call or SMS alarm complete pipeline network leak.
Detect pipe network leakage failure warning system example as Figure 10 illustrates one, comprise a heat distribution pipe network system 1001; A data acquisition system (DAS), wherein data acquisition system (DAS) comprise circulation pump of heat-supply network 1003, user side import and export 1004 and install wireless pressure sensor 1005 form; A communication system 1006; A heating network operation management system 1002; A warning system 1008.Its workflow is: after the import and export pressure data of water circulating pump 1003 and user side 1004 is gathered by wireless pressure sensor 1005, heat supply network management system 1002 is sent to by wired or wireless communication network 1006, in heat supply network management system 1002, Real-Time Monitoring is carried out to pipe network, and carry out leakage failure testing by the method for recognizer, obtain recognition result.Signal to be changed by the communication converter 1007 that is connected to computer end and is transferred to alarm controller by recognition result, finally outputed signal by alarm controller, sound and light alarm, telephone call or SMS alarm are given the alarm, completes the realization of system of the present invention.
Claims (5)
1. whether one kind exist the recognition methods of the data-driven of leakage for detecting heating network; It is characterized in that the import and export pressure Real-time Monitoring Data according to water circulating pump or user side, data are carried out certain conversion according to the input requirements of recognizer, pipe network real-time running state when identifying normal operation or leakage failure occurs, carries out identifying, detecting.
2. the method for claim 1, is characterized in that pipe network real-time running state step during leakage failure: search on the time-serial position of pressure data in chronological order; The characteristic and recognition rule of each point on curve or each section is mated; Determine real-time status; If its feature meets the leakage failure feature in recognition rule, then think that this section there occurs leakage failure; Otherwise, determine that this section is normal operating condition.
3. method as claimed in claim 2, is characterized in that the process of establishing of leaking recognition rule is: first, the Real-time Monitoring Data of pipe network is collected; After data prediction, be used to that algorithm is described to curve and train, verify and test process; Finally export the recognition rule of leakage failure.
4. method as claimed in claim 3, is characterized in that curve describes algorithm and carries out training, to verify and test process is: be divided into training set and test set in the data that pre-service is good; Subsequently, training set is used to training and the checking of recognizer.
5. one kind is detected the warning system of heating network leakage failure based on the recognizer by data-driven, comprise basic unit's heat distribution pipe network system, data acquisition system (DAS) and communication system, pipe network operation management system, warning system, heat distribution pipe network system connects tensimeter or the wireless pressure sensor of water circulating pump and user side, the data of collection are that medium is sent to pipe network operation management system with wireless network by data acquisition system (DAS) connecting communication system, pipe network operation management system is through carrying out the pre-service of data, the extraction of feature and the realization of recognizer, identify, signal to be changed by the communication converter that is connected to computer end and is transferred to alarm controller by recognition result, finally outputed signal by alarm controller, make sound and light alarm, telephone call or SMS alarm give the alarm, complete the realization of system of the present invention.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510749968.9A CN105388884A (en) | 2015-11-05 | 2015-11-05 | Alarm system for detecting leakage fault of heat supply network based on identification algorithm driven by data and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510749968.9A CN105388884A (en) | 2015-11-05 | 2015-11-05 | Alarm system for detecting leakage fault of heat supply network based on identification algorithm driven by data and method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105388884A true CN105388884A (en) | 2016-03-09 |
Family
ID=55421258
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510749968.9A Pending CN105388884A (en) | 2015-11-05 | 2015-11-05 | Alarm system for detecting leakage fault of heat supply network based on identification algorithm driven by data and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105388884A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108416106A (en) * | 2018-02-05 | 2018-08-17 | 江苏方天电力技术有限公司 | Feed pump fault detection method based on multiscale principal component analysis |
| CN109885951A (en) * | 2019-02-28 | 2019-06-14 | 中科云创(厦门)科技有限公司 | Equipment fault diagnosis method and device |
| CN110779745A (en) * | 2019-10-12 | 2020-02-11 | 杭州安脉盛智能技术有限公司 | Heat exchanger early fault diagnosis method based on BP neural network |
| CN112990059A (en) * | 2021-03-30 | 2021-06-18 | 安徽天能清洁能源科技有限公司 | IV curve diagnosis method and device based on curve shape description |
| CN113486950A (en) * | 2021-07-05 | 2021-10-08 | 华能国际电力股份有限公司上安电厂 | Intelligent pipe network water leakage detection method and system |
| CN113657019A (en) * | 2021-07-06 | 2021-11-16 | 大唐互联科技(武汉)有限公司 | Heat supply pipe network early warning system |
| CN113685736A (en) * | 2021-08-24 | 2021-11-23 | 上海观燃智能科技有限公司 | Method and system for detecting leakage of gas pipeline network based on pressure parameter analysis |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1367374A (en) * | 2002-01-24 | 2002-09-04 | 天津大学 | Heat-supply network failure quick detection system |
| CN101255950A (en) * | 2007-06-14 | 2008-09-03 | 北京华银广誉石油科技有限公司 | Intelligent warning system for computation and pipe monitoring of oil transportation collection and transmission system |
| CN102174994A (en) * | 2011-03-11 | 2011-09-07 | 天津大学 | Pipe burst accident on-line positioning system for urban water supply pipeline network |
| CN102865459A (en) * | 2012-09-26 | 2013-01-09 | 华北电力大学（保定） | Heat supply pipe network leakage positioning system and heat supply pipe network leakage positioning method |
| CN203023812U (en) * | 2013-01-16 | 2013-06-26 | 苏州市职业大学 | Oil pipeline leakage monitoring system based on wireless sensing network |
| WO2015123916A1 (en) * | 2014-02-19 | 2015-08-27 | 清华大学深圳研究生院 | Method for measuring and scheduling node flow of pipe network based on pressure monitoring |
-
2015
- 2015-11-05 CN CN201510749968.9A patent/CN105388884A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1367374A (en) * | 2002-01-24 | 2002-09-04 | 天津大学 | Heat-supply network failure quick detection system |
| CN101255950A (en) * | 2007-06-14 | 2008-09-03 | 北京华银广誉石油科技有限公司 | Intelligent warning system for computation and pipe monitoring of oil transportation collection and transmission system |
| CN102174994A (en) * | 2011-03-11 | 2011-09-07 | 天津大学 | Pipe burst accident on-line positioning system for urban water supply pipeline network |
| CN102865459A (en) * | 2012-09-26 | 2013-01-09 | 华北电力大学（保定） | Heat supply pipe network leakage positioning system and heat supply pipe network leakage positioning method |
| CN203023812U (en) * | 2013-01-16 | 2013-06-26 | 苏州市职业大学 | Oil pipeline leakage monitoring system based on wireless sensing network |
| WO2015123916A1 (en) * | 2014-02-19 | 2015-08-27 | 清华大学深圳研究生院 | Method for measuring and scheduling node flow of pipe network based on pressure monitoring |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108416106A (en) * | 2018-02-05 | 2018-08-17 | 江苏方天电力技术有限公司 | Feed pump fault detection method based on multiscale principal component analysis |
| CN108416106B (en) * | 2018-02-05 | 2022-02-08 | 江苏方天电力技术有限公司 | Water feeding pump fault detection method based on multi-scale principal component analysis |
| CN109885951A (en) * | 2019-02-28 | 2019-06-14 | 中科云创(厦门)科技有限公司 | Equipment fault diagnosis method and device |
| CN110779745A (en) * | 2019-10-12 | 2020-02-11 | 杭州安脉盛智能技术有限公司 | Heat exchanger early fault diagnosis method based on BP neural network |
| CN110779745B (en) * | 2019-10-12 | 2021-07-06 | 杭州安脉盛智能技术有限公司 | Heat exchanger early fault diagnosis method based on BP neural network |
| CN112990059A (en) * | 2021-03-30 | 2021-06-18 | 安徽天能清洁能源科技有限公司 | IV curve diagnosis method and device based on curve shape description |
| CN113486950A (en) * | 2021-07-05 | 2021-10-08 | 华能国际电力股份有限公司上安电厂 | Intelligent pipe network water leakage detection method and system |
| CN113486950B (en) * | 2021-07-05 | 2023-06-16 | 华能国际电力股份有限公司上安电厂 | Intelligent pipe network water leakage detection method and system |
| CN113657019A (en) * | 2021-07-06 | 2021-11-16 | 大唐互联科技(武汉)有限公司 | Heat supply pipe network early warning system |
| CN113685736A (en) * | 2021-08-24 | 2021-11-23 | 上海观燃智能科技有限公司 | Method and system for detecting leakage of gas pipeline network based on pressure parameter analysis |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105388884A (en) | Alarm system for detecting leakage fault of heat supply network based on identification algorithm driven by data and method | |
| CN109555979B (en) | Water supply pipe network leakage monitoring method | |
| CN107941537B (en) | A kind of mechanical equipment health state evaluation method | |
| CN106761668B (en) | Oil well failure intelligent analysis decision system and method | |
| CN105846780B (en) | A kind of photovoltaic module method for diagnosing faults based on decision-tree model | |
| Liu et al. | A leak detection method for oil pipeline based on Markov feature and two-stage decision scheme | |
| CN111311059B (en) | Waterwheel house fault diagnosis method based on knowledge graph | |
| CN104390657B (en) | A kind of Generator Unit Operating Parameters measurement sensor fault diagnosis method and system | |
| CN109854953B (en) | Crude oil conveying pipeline leakage detection system and method under special working condition | |
| CN201943686U (en) | Real-time early warning system for oil well fault | |
| CN108981796B (en) | Five-in-one hydraulic mechanical fault diagnosis method | |
| CN110285330B (en) | Water network pipe burst detection method based on local outlier factor | |
| CN116658492B (en) | Intelligent power catwalk and method thereof | |
| CN109492790A (en) | Wind turbines health control method based on neural network and data mining | |
| KR102295879B1 (en) | Method and System for Real-time Leakage State Recognition of Tubes | |
| Zhou et al. | Structural health monitoring of offshore wind power structures based on genetic algorithm optimization and uncertain analytic hierarchy process | |
| CN113486950B (en) | Intelligent pipe network water leakage detection method and system | |
| CN115614292B (en) | Vibration monitoring device and method for vertical water pump unit | |
| CN117560300B (en) | Intelligent internet of things flow prediction and optimization system | |
| CN102242872B (en) | Oil transportation pipeline network leakage detection method based on generalized fuzzy hyperbolic model | |
| CN103617105A (en) | Self-adaptation multilevel flow model equipment diagnosis method based on data driving | |
| CN116090347A (en) | Intelligent monitoring and early warning system for historical building structure under steady load | |
| KR20230137021A (en) | detecting method and system of pipe abnormality using artificial intelligence | |
| CN114442543A (en) | Computer monitoring method suitable for early warning of hydropower station fault | |
| KR102602840B1 (en) | Training dataset generating system and training dataset generating method using the same |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20160309 |
|
| WD01 | Invention patent application deemed withdrawn after publication |