CN101719315B - Method for acquiring dynamic traffic information based on middleware - Google Patents
Method for acquiring dynamic traffic information based on middleware Download PDFInfo
- Publication number
- CN101719315B CN101719315B CN2009102565373A CN200910256537A CN101719315B CN 101719315 B CN101719315 B CN 101719315B CN 2009102565373 A CN2009102565373 A CN 2009102565373A CN 200910256537 A CN200910256537 A CN 200910256537A CN 101719315 B CN101719315 B CN 101719315B
- Authority
- CN
- China
- Prior art keywords
- data
- information
- traffic
- road
- checkout equipment
- 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.)
- Expired - Fee Related
Links
Images
Landscapes
- Traffic Control Systems (AREA)
Abstract
The invention relates to a method for acquiring dynamic traffic information based on a middleware, which overcomes the shortcomings of data loss, data noise, and particularly the multi-source isomerism of data, reduces redundant data, ensures the accuracy of the data and improves the accuracy and the reliability of the data. The method comprises the following steps of: 1) transmitting the traffic information by adopting a serial interface communication mode and/or a network communication mode; 2) customizing an information acquisition port which can be matched with different inspection devices of the traffic information by using an interface definition language IDL in the CORBA middleware technique, identifying and normalizing the data from different inspection devices, thus acquiring the real-time dynamic traffic information, such as road traffic flow, vehicle velocity, road occupancy rate and the like; 3) performing preprocessing on the all data acquired by the method, and performing map matching on a floating automobile by using a road matching algorithm based on a network topology relationship; and 4) fusing and saving the multi-source isomerous real-time dynamic traffic data which is preprocessed by the method in a database by using an immune clustering neural network.
Description
Technical field
Patent of the present invention is a kind of method for acquiring dynamic traffic information based on middleware, belongs to traffic information collection and processing technology field.
Background technology
Along with socioeconomic continuous development and quickening of urbanization process, urban population and motor vehicles increase day by day, the problems such as environmental pollution that the traffic congestion of bringing thus, traffic hazard and energy dissipation and the automobile emission that is multiplied cause, not only seriously restrict city and continuous development of society economy, also seriously affecting city dweller's quality of life simultaneously.
Intelligent transportation is to solve modern society's transport need and one of important channel of supplying with contradiction, and the transport information platform is the core content of intelligent transportation system, is the hinge of realizing message exchange and information sharing between each subsystem of intelligent transportation.For the transport information and other relevant informations that insert from each subsystem of intelligent transportation system and traffic relevant departments, the transport information platform adopts advanced data storage technology that it is stored and manages, and use various advanced persons' intelligence computation and data fusion technology that it is carried out analyzing and processing, for each subsystem of intelligent transportation system provides timely, accurate, full and accurate data message, for traffic relevant departments provide the data information sharing service, for vast traffic participant provides based on information service in the system-wide and aid decision making.
The urban dynamic traffic information platform is made up of traffic information collection subsystem, information processing and analyzing subsystem, information issue subsystem, data storage and several major parts of data base management subsystem.The collection of transport information is the basis that data processing and informix utilize, and is technical difficulty maximum in the urban transportation Study of intelligent, also is one of core technology of being badly in need of most.Based on the optimum management of transport information Platform Implementation urban highway traffic and the foundation and the key foundation of control is to obtain and fully utilize the different kinds of roads transport information, particularly about dynamic informations such as road traffic flow, car speed and roadway occupancies.These information mainly obtain by various fixed traffic information collection equipment (as coil checker, microwave radar detecting device, video detector etc.) and Floating Car mobile collection equipment at present.
Aspect the transmission of transport information, all optional ethernet interface of joining of main flow equipment of urban road traffic information detection at present, part transport information checkout equipment also can provide the RS232/422/485 serial communication interface.In addition, development along with computer networking technology, network transmission scheme based on Internet has obtained widespread use in the novel traffic infosystem, adopt at present the more dual mode that has: a kind of is serial server and data optical receiver mode closely, and another kind is the whole network transmission mode of Ethernet fiber optic.No matter be serial server or network fiber optic transmission mode, all but the direction towards management type develops, selected equipment all can be included in the unified network management scope, and this just requires the network fiber optic of front end and serial server need possess network management function.In addition, the wireless GPRS transmission also is one of important communication mode of transport information, and GPRS modulator-demodular unit transmission plan is laid convenient, is particularly suitable for the occasion that original data transmission network can't use in the traffic system improvement project.The floating car traffic information collection is that the typical case of wireless GPRS transmission plan uses.These three kinds of schemes of serial server, network fiber optic and wireless GPRS all are based on the network communication mode of IP address.
Along with deepening continuously of traffic information collection technical research, the multi-sensor information of mutual supplement with each other's advantages detects the main direction that has become the traffic information collection technical development.The Real-time Road traffic data that inserts the transport information platform comes from various transport information checkout equipments or the artificial information acquisition source that is distributed on the city road network.The Traffic Information checkout equipment of robotization comprises fixed checkout equipment and Floating Car mobile detection apparatus such as coil checker, microwave detector, video detector.Data fixed-site, scope that fixed checkout equipment is gathered are narrow and small relatively, but have image data characteristics accurately and reliably.The Floating Car collecting device, acquisition range is extensive, data volume is big owing to have, and has become one of important means of present urban transportation information detection.But owing to city overhead road or other skyscrapers generally do not provide the common influence of inner reasons such as complicated algorithm to external causes such as stopping of GPS positioning signal and mobile unit, the data accuracy that the Floating Car collecting device collects is generally lower.Various transport information checkout equipments respectively have relative merits, can detected traffic parameter kind and form, particularly the data layout of traffic parameter may have nothing in common with each other, cause traffic data to have typical polyphyly thus, isomerism, multi-stratification, imperfection and inconsistency, and has a time and space feature, data processing techniques such as necessary employing information fusion improve the reliability and the accuracy of transport information, with out of true, imperfect, inconsistent, unreliable, even conflicting transport information changes into target or conforming explanation of phenomenon and description.
The multi-source heterogeneous transport information source that the plurality of communication schemes of considering the acquisition of road traffic information system coexists, multiple transport information checkout equipment provides and the complex process of transport information, be the structure of simplifying traffic information acquisition system, the data layout of standard transport information, so that the post-processed of transport information and comprehensive utilization, this patent has been invented a kind of method for acquiring dynamic traffic information based on middleware, and its system architecture and data communication mode are as shown in Figure 1.
This method utilizes middleware Technology to carry out unified organizational system and specification handles from the data layout of different transport information checkout equipments, and pre-service and data fusion by data, overcome the harmful effect that the factor such as multi-source heterogeneous of loss of data, data noise, particularly data causes whole traffic information system, reduced redundant data, guarantee the accuracy of data, improved the accuracy and the reliability of data.
Summary of the invention
The present invention is directed to the multiple transport information checkout equipment that generally uses in the present municipal intelligent traffic system, and corresponding multi-source heterogeneous dynamic information pattern, a kind of method for acquiring dynamic traffic information provided based on middleware Technology.This method will be carried out unified organizational system and specification handles from the data layout of different transport information checkout equipments, adopt serial communication and/or network communication mode to realize the transmission of the multi-source heterogeneous multidate information of urban highway traffic, utilize abnormal data reparation and missing data filling technique to the data pre-service, and the traffic data of Floating Car collection carried out map match, utilize the immune cluster neural network that the multi-source heterogeneous real-time dynamic information from different checkout equipments is carried out fusion treatment at last, and deposit database in.
For achieving the above object, the present invention has adopted following technical scheme:
1) adopt serial interface communication mode and/or network communication mode to carry out the transmission of transport information;
2) the information acquisition port that utilizes the IDL (Interface Definition Language) IDL customization in the CORBA middleware Technology to be complementary with different transport information checkout equipments, identification and the data of standard from different checkout equipments realize the collection to real-time dynamic informations such as road traffic flow, car speed and roadway occupancies;
3) all data that collect are carried out pre-service, and adopt the road matching algorithm of topological relation Network Based to carry out the map match of Floating Car;
4) utilize the immune cluster neural network, pretreated multi-source heterogeneous real-time dynamic traffic data are merged and deposit database in.
Described step 2) utilize the IDL (Interface Definition Language) IDL in the CORBA middleware Technology to carry out unified organizational system in from the data layout of different transport information checkout equipments, identification and the data of standard from different checkout equipments, with the convenient dynamic road transport information of handling with multi-source heterogeneous feature, collection port and all kinds of transport information checkout equipment are complementary, gather real-time dynamic informations such as road traffic flow, car speed and roadway occupancy from different transport information checkout equipments.
Wherein, to doing following unified Definition from the data layout of different transport information checkout equipments: the definition event structure comprises incident sequence number and event description; The checkout equipment type comprises coil checker, microwave detector, video detector and Floating Car checkout equipment; The external communication protocol that checkout equipment adopts comprises RS-232/422/485,802.1, GPRS and Ethernet; The essential information of checkout equipment comprises: brand, manufacturer, unit type, device type, device numbering and communication protocol; Discharge pattern comprises forward flow and reverse flow; Speed type comprises forward direction speed and inverted speed; The occupation rate type comprises forward occupation rate and reverse occupation rate; The checkout equipment state comprise success, invalid, finish and do not finish; The checkout equipment acquisition parameter comprises flow information, velocity information and occupation rate information; The checkout equipment operation comprises job title and job description.
To the following operation of information acquisition port definition: the checkout equipment initialization, and be connected with collection terminal; Judge whether checkout equipment is in active state; Disconnect and connecting; Suspend; Get back to initial bit; Upload operation; Download operation; Gather the road traffic flow; Collection vehicle speed; Gather roadway occupancy.
In the described step 3), the pre-service of data is comprised the reparation and the missing value estimation of abnormal data.Wherein, the data exception value adopts misdata to limit and handles, passing threshold method identification error data, promptly set corresponding data threshold according to data type, the threshold ratio up and down of traffic parameter such as flow, speed and the occupation rate that checkout equipment is gathered and its setting, if measured value is then thought misdata not in the scope of last lower threshold value defined, it is repaired.
The threshold value of each traffic parameter is determined according to the planning index of city road networks such as the specified traffic capacity of road and the historical statistical data of each traffic parameter.Abnormal data is according to following formula reparation:
q
mod(t)＝αq
lw(t)+(1-α)q(t-1)
Wherein: q
Mod(t) be the abnormal data reparation value of traffic parameter q (t); q
Lw(t) be the measured value of last one all same working days, same period traffic parameter q (t); Q (t-1) was the measured value of a last period of traffic parameter q (t); The weights of α ∈ [0,1] for calculating, if certain traffic parameter has bigger time jitter characteristic, α gets less numerical value, otherwise α gets bigger numerical value.
The disappearance of data brings adverse effect for equally the further utilization of data.The recognition methods of missing data be the data definition that obtains in the certain hour section is become a certain moment data (for example, the time interval of regulation image data is 5 minutes, then the data that obtain in this time period to 9:05 at 9:00 all are considered as the data of 9:00), then the time period of data is scanned, if do not obtaining data in the section sometime, think then that the data of this period have produced to lose.
The missing value estimation value is taken as the weighted sum of the measured value of a period on the measured value of all same working days, same period on the same traffic parameter and this traffic parameter, and computing formula is identical with the reparation of abnormal data.
Because urban road traffic flow has characteristic seven days workaday quasi-periodicitys, and traffic behavior has the continuous-changing features in certain period, said method can be realized reparation or the missing value estimation for unusual traffic data more exactly, satisfy the quality of data requirement that urban road traffic information is handled, and calculate simply, and be easy to Project Realization.
In the described step 3), map-matching method is: the vehicle location information that GPS is received is planned for map coordinates system, determine vehicle initial position road ID number, according to initial road ID number, determine the spatial data and the attribute data of each node on start node numbering, terminal node numbering and this road of this road; Calculate the GPS locating information (vehicle is promptly located at that point for X, Y) nearest point to this road, and the judgment criterion that adopts when calculating is:
Wherein (X Y) is GPS locating information, (X
i, Y
i) be the coordinate points in the road.
In the described step 4),, be a kind of process of utilizing radial basis function neural network that pretreated multi-source heterogeneous real-time dynamic traffic data are merged based on the data fusion technology of immune cluster neural network.Wherein the hidden layer structure of radial basis function neural network (mainly being the number of hidden node and the center of basis function) is the key factor of decision network approximation capability, the hidden layer central point that empirical method is manually set number, the picked at random basis function of hidden node all is difficult to guarantee the network structure that obtains to optimize, and then influences the data fusion speed and the accuracy of network.The base this, the present invention is based on the Immune Clone Selection and the immunological memory mechanism of Immune System, provided a kind of improved artificial immunity data clusters new method, by immune cluster to historical traffic data collection, the hidden layer structure of optimization network, to obtain the network integration model of optimum structure, realize the efficient fusion of transport information.
Beneficial effect of the present invention is: the method for acquiring dynamic traffic information based on middleware of proposition, thinking is novel and be easy to realize.It utilizes the IDL (Interface Definition Language) IDL of CORBA middleware Technology to carry out unified organizational system from the multi-source heterogeneous data of different transport information checkout equipments, realized the standardization processing of transport information checkout equipment data layout, simplified the interface of traffic information database and various transport information checkout equipments, and can be applied to other collection of multi-sensor informations with multi-source heterogeneous feature fields, have wide range of applications; By the mutual supplement with each other's advantages of different transport information checkout equipments, realized quality data collection for real-time dynamic informations such as road traffic flow, car speed and roadway occupancies; Integration by serial communication and two kinds of communication pattern traffic information collections of network service mode, make acquisition means diversified more, thereby can satisfy the information acquisition of different levels road transport information and the objective requirement of data transmission and processing better; By to the pre-service of data with based on the data fusion technology of immune cluster neural network, overcome the harmful effect that loss of data in data transmission procedure, data noise, particularly factor such as multi-source heterogeneous cause total system, also reduced simultaneously a large amount of redundant datas, guarantee the accuracy of data, and significantly improved the accuracy and the reliability of data.
Description of drawings
Fig. 1 entire system Organization Chart;
Fig. 2 traffic data collection traffic diagram;
Fig. 3 serial communication bottom class method process flow diagram;
Fig. 4 network communication program process flow diagram;
Fig. 5 radial basis function neural network structural drawing;
Fig. 6 immune cluster neural metwork training process flow diagram.
Embodiment
Below the specific embodiment of the present invention is described in detail.
A kind of method for acquiring dynamic traffic information based on middleware, its step is:
1) adopt serial interface communication mode and/or network communication mode to carry out the transmission of transport information;
2) the information acquisition port that utilizes the IDL (Interface Definition Language) IDL customization in the CORBA middleware Technology to be complementary with different transport information checkout equipments, identification and the data of standard from different checkout equipments realize the collection to real-time dynamic informations such as road traffic flow, car speed and roadway occupancies;
3) all data that collect are carried out pre-service, and adopt the road matching algorithm of topological relation Network Based to carry out the map match of Floating Car;
4) utilize the immune cluster neural network, pretreated multi-source heterogeneous real-time dynamic information data are carried out data fusion and deposited database in.
The present invention is as a kind of dynamic road traffic information collection technology based on middleware, its system architecture and data communication mode as shown in Figure 1, comprise the integration of definition, serial communication and two kinds of data-transmission modes of network service of image data form, based on the uniform data form of customization to discern pre-service and the data fusion with standardization, data from the data of different acquisition equipment.
At first, middleware is a kind of pattern of setting up data application resource interoperability in distributed system, realizes that database under the multi-source heterogeneous environment connects or the connection of file system.The Data-collection middleware interactive operation between service layer and the data Layer that is used to manage business, purpose is that the complicacy of collecting device and data access is isolated, can avoid writing of abnormal data at collecting part, can avoid applications in the processing section directly to the read-write of data-base content and the maloperation that causes.
The principal feature of Data-collection middleware design be with being connected and visiting effectively and manage of database, by management to data connection and access mechanism, improve multi-user access performance of database on the network, optimize Network Transmission, and being connected of support and several data storehouse.
For making Data-collection middleware have better generality, the present invention will carry out unified organizational system by IDL from the data of different transport information checkout equipments, collection port and all kinds of transport information checkout equipment are complementary, the standardization processing of data layout and the mutual supplement with each other's advantages of multiple checkout equipment have been realized, simplified the interface of database and various checkout equipments, realized easily for the road traffic flow, the high-quality collection of Real-time and Dynamic such as car speed and roadway occupancy Traffic Information is also guaranteed simultaneously can gather and be sent to traffic information database smoothly to the transport information that the checkout equipment that dissimilar transport information checkout equipments or system increase newly provides.By these interfaces are integrated in the idl file, simplified detail code, assembly is concentrated more finished integrated functionality.
In the interface of integrated idl file is realized, use IDL (Interface Definition Language) IDL to do following unified Definition to the data layout from different transport information checkout equipments: the definition event structure comprises incident sequence number and event description; The checkout equipment type comprises coil checker, microwave detector, video detector and Floating Car checkout equipment; The external communication protocol that checkout equipment adopts comprises RS-232/422/485,802.1, GPRS and Ethernet; The essential information of checkout equipment comprises: brand, manufacturer, unit type, device type, device numbering and communication protocol; Discharge pattern comprises forward flow and reverse flow; Speed type comprises forward direction speed and inverted speed; The occupation rate type comprises forward occupation rate and reverse occupation rate; The checkout equipment state comprise success, invalid, finish and do not finish; The checkout equipment acquisition parameter comprises flow information, velocity information and occupation rate information; The checkout equipment operation comprises job title and job description.
To the following operation of information acquisition port definition: the checkout equipment initialization, and be connected with collection terminal; Judge whether checkout equipment is in active state; Disconnect and connecting; Suspend; Get back to initial bit; Upload operation; Download operation; Gather the road traffic flow; Collection vehicle speed; Gather roadway occupancy.
Secondly, introduce the performing step of the traffic information collection method based on middleware of the present invention in detail:
Step 1:
The all optional ethernet interface of joining of main flow equipment of urban road traffic information detection at present, part transport information checkout equipment also provides the RS232/422/485 serial communication interface.These transport information checkout equipments generally are laid in and respectively detect highway section or intersection, with communicating by letter according to the networking situation between the traffic information center, can adopt various communication mode.
Fixed wagon detector distance communication station nearby is hundreds of rice at least, and several at most kms if directly the communication interface that provides of employing equipment is carried out data transmission, obviously can't satisfy the required distance of data transmission.The conventional traffic information detecting system generally adopts detecting device to connect modulator-demodular unit (Modem) and inserts public telephone network (PSTN), connects Modem again to monitoring client, imports data into the communication server by the multi-channel serial port card.This is a kind of serial communication mode.
Development along with computer networking technology, network transmission scheme based on Internet has obtained extensive application, adopt at present the more dual mode that has: a kind of is serial server and data optical receiver mode closely, and another kind is the whole network transmission mode of Ethernet fiber optic.No matter be serial server or network fiber optic transmission mode, all but the direction towards management type develops, selected equipment all can be included in the unified network management scope, and this just requires the network fiber optic of front end and serial server must possess network management function.In addition, the wireless GPRS transmission also is one of important communication mode of transport information transmission.GPRS modulator-demodular unit transmission plan is laid convenient, is particularly suitable for the occasion that original data transmission network can't use in the traffic system improvement project.Along with constantly improving and the progressively decline of rate of GPRS network, its range of application also will further be enlarged.
These three kinds of schemes of serial server, network fiber optic and wireless GPRS all are based on the network communication mode of IP address.The wireless GPRS transmission plan is mainly adopted in the floating car traffic information collection, also is a kind of IP address-based network communication mode.
The traffic data collection communication pattern as shown in Figure 2.
The traffic surveillance and control center communication server will be gathered real-time dynamic information based on serial communication and/or network service dual mode in view of the above.The concrete communication steps of two kinds of communication modes is as described below:
(1) serial interface communication mode
The 1-1 serial communication cycle begins;
1-2 opens serial ports, and messaging parameter is set;
1-3 needs the detection of address, if detect this address, then determines address list; If do not detect, then use the default address tabulation;
1-4 carries out Data Receiving;
1-5 sends read/write communication request packet to collection terminal;
1-6 receives the collection terminal packet;
Whether 1-7 judgment data packet format correct, and if correctly would carry out data presentation and the storage, change step 1-4 then over to; If incorrect, then return step 1-5.
The flow process of serial interface communication mode as shown in Figure 3.
(2) network communication mode
The 2-1 network service cycle begins;
2-2 opens port, and messaging parameter is set;
2-3 needs the detection of address; If detect, then determine address list; If do not detect, then use the default address tabulation;
2-4 carries out Data Receiving;
2-5 sends read/write communication request packet to collection terminal;
2-6 receives the collection terminal packet;
Whether 2-7 judgment data packet format is correct, if correctly then change step 2-4 over to; If incorrect, then change step 2-5 over to.
The flow process of network communication mode as shown in Figure 4.
Step 2:
Because the urban transportation intelligent development is asynchronous, the transport information checkout equipment that the different highway sections of urban road network or traffic zone adopt may be mutually different, the for example early stage ground induction coil detecting device that adopts more, then adopt fixed transport information checkout equipments such as microwave detector, video detector now more, or Floating Car dynamic information checkout equipment.Along with improving constantly of urban transportation intelligent management level, accuracy, accuracy and real-time to transport information have proposed more and more higher requirement, same place just having occurred in the transport information context of detection may coexist by multiple checkout equipment, detects application state with the multisensor transport information that realizes having complementary advantages.Because the data layout of different checkout equipments may be mutually different, and then causes the detection data of Traffic Information to have typical multi-source heterogeneous feature.
For making Data-collection middleware have better generality, the present invention will carry out unified organizational system by IDL (Interface Definition Language) IDL from the data layout of different transport information checkout equipments, collection port and each transport information checkout equipment are complementary, the mutual supplement with each other's advantages of multiple checkout equipment and the standardization processing of data layout have been realized, simplified the interface of database and various checkout equipments, realized easily for the road traffic flow, the high-quality collection of Real-time and Dynamic such as car speed and roadway occupancy Traffic Information is also guaranteed simultaneously can gather and be sent to traffic information database smoothly to the transport information that the checkout equipment that dissimilar transport information checkout equipments or system increase newly provides.By these interfaces are integrated in the idl file, simplified detail code, assembly is concentrated more finished integrated functionality.
In the interface of integrated idl file is realized, to having done following unified Definition from the data layout of different transport information checkout equipments: the definition event structure comprises incident sequence number and event description; The checkout equipment type comprises coil checker, microwave detector, video detector and Floating Car checkout equipment; The external communication protocol that checkout equipment adopts comprises RS-232/422/485,802.1, GPRS and Ethernet; The essential information of checkout equipment comprises: brand, manufacturer, unit type, device type, device numbering and communication protocol; Discharge pattern comprises forward flow and reverse flow; Speed type comprises forward direction speed and inverted speed; The occupation rate type comprises forward occupation rate and reverse occupation rate; The checkout equipment state comprise success, invalid, finish and do not finish; The checkout equipment acquisition parameter comprises flow information, velocity information and occupation rate information; The checkout equipment operation comprises job title and job description.
To the following operation of information acquisition port definition: the checkout equipment initialization, and be connected with collection terminal; Judge whether checkout equipment is in active state; Disconnect and connecting; Suspend; Get back to initial bit; Upload operation; Download operation; Gather the road traffic flow; Collection vehicle speed; Gather roadway occupancy.
Uniform data form based on above-mentioned customization, at first data attributes such as the source of the data that collect, type are discerned, and, make its uniform formatization, so that carry out follow-up data fusion, storage and processing and utilization to the form of these data specification handles in addition.
Step 3:
The dynamic road traffic information system gathered and uploaded to information center in real time by all kinds of checkout equipments raw data is not all to be correct complete, need carry out the pre-service such as cleaning of data.In addition, Traffic Informations such as the real-time positioning of Floating Car collection and speed are uploaded to information center and do not mate physical location fully, need carry out map match, to guarantee the accuracy and the validity of locating information.
The Real-time Road traffic data of traffic information center comes from various transport information checkout equipments or the artificial information acquisition source that is distributed on the city road network.The Traffic Information checkout equipment of robotization mainly comprises fixed coil checker, microwave detector, video detector and Floating Car mobile detection apparatus.Various checkout equipments respectively have relative merits; can detected traffic parameter (as the magnitude of traffic flow, car speed, time headway, vehicle classification, lane occupancy ratio, queue length etc.) kind and form and data layout may have nothing in common with each other; and because the existence of multiple factors such as Acquisition Error, equipment failure; the raw data of checkout equipment output may exist the mistake of data, unusual or loss of data; cause the imperfect and/or inconsistent of data, therefore at first need the data of each data source are carried out correctness, integrality and conforming check.Owing to loss of datas that reason caused such as weather conditions or communication system failure, also need to adopt certain technical method and means they to be repaired or provide alternate data in addition.
Incomplete statistics, in the transport information detection system, misdata often accounts for 25% of data total amount, and obliterated data can reach 15% of total amount of data, and therefore, the data pre-service is the requisite information processing function of traffic information acquisition system.
The data preconditioning technique that the present invention relates to mainly comprises for contents such as the repairing of abnormal data in the original detection data and missing value estimation.
The exceptional value of data (being commonly called as bad value) is meant with the objective condition of measuring and can not be interpreted as reasonably, obviously departs from the overall indivedual measured values of measurement.Exceptional value can directly influence the overall correctness of data.In data acquisition system (DAS), the main cause that exceptional value occurs generally is that checkout equipment fault and probability of occurrence are minimum but act on extremely strong factors such as sporadic interference.
For above-mentioned abnormal data, the present invention adopts misdata to limit and handles, passing threshold method identification error data, promptly set corresponding data threshold according to data type, the threshold ratio up and down of traffic parameter such as flow, speed and the occupation rate that checkout equipment is gathered and its setting, if measured value is then thought misdata not in the scope of last lower threshold value defined, it is repaired.
The threshold value of each traffic parameter is determined according to the planning index of city road networks such as the specified traffic capacity of road and the historical statistical data of each traffic parameter.Abnormal data is according to following formula reparation:
q
mod(t)＝αq
lw(t)+(1-α)q(t-1)
Wherein: q
Mod(t) be the abnormal data reparation value of traffic parameter q (t); q
Lw(t) be the measured value of last one all same working days, same period traffic parameter q (t); Q (t-1) was the measured value of a last period of traffic parameter q (t); The weights of α ∈ [0,1] for calculating, if certain traffic parameter has bigger time jitter characteristic, α gets less numerical value, otherwise α gets bigger numerical value.
The disappearance of data brings adverse effect for equally the further utilization of data.The present invention for the recognition methods of missing data be the data definition that obtains in the certain hour section is become a certain moment data (for example, the time interval of regulation image data is 5 minutes, then the data that obtain in this time period to 9:05 at 9:00 all are considered as the data of 9:00), then the time period of data is scanned, if do not obtaining data in the section sometime, think then that the data of this period have produced to lose.
The missing value estimation value is taken as the weighted sum of the measured value of a period on the measured value of all same working days, same period on the same traffic parameter and this traffic parameter, and computing formula is identical with the reparation of abnormal data.
Because urban road traffic flow has characteristic seven days workaday quasi-periodicitys, and traffic behavior has the continuous-changing features in certain period, said method can be realized reparation or the missing value estimation for unusual traffic data more exactly, satisfy the quality of data requirement that urban road traffic information is handled, and calculate simply, and be easy to Project Realization.
In the gatherer process of floating car traffic information, owing to be subjected to the influence of GPS bearing accuracy, and city overhead road or other skyscrapers cause the GPS locating information generally to have bigger deviation to reasons such as stopping of GPS positioning signal, even constantly may lose locating information fully at some.Therefore, if directly the GPS locating information is loaded on the electronic chart, anchor point will depart from real road, and the traffic data that causes collecting can not accurately reflect the true traffic of road network.On the other hand, because the restriction of cost, the mobile unit of Floating Car generally can not carry out complex calculation such as real time differential for the gps satellite positioning signal that receives and realize high-precision location, at this moment position the calibration of signal with regard to the map matching technology that needs employing information center, so processing also can utilize the road information of degree of precision to revise the error of positioning system, to improve the performance of total system.
Map match (Map Matching) is a kind of location modification method based on software engineering, its basic thought is that the vehicle location track of locating device acquisition and the road information in the electronic map database are interrelated, determine the particular location that vehicle is positioned at map thus, and with positioning track relatively, determine the most possible running section of vehicle and vehicle most possible position in this highway section by suitable matching process with road information.
Map-matching algorithm can be divided into two relatively independent processes: the one, and the road of searching vehicle current driving, i.e. all road sections of search and combination thereof in the adjacent domain of vehicle flight path, ask for the matching degree value of the combination of these roads and vehicle flight path respectively, with the road of optimum matching current driving road as vehicle; The 2nd, current anchor point is navigated on the road of vehicle current driving.
Logical theoretical research and traffic practice, the present invention adopts the road matching algorithm of topological relation Network Based.Main points are as follows: at first the vehicle location information that GPS is received is planned among the map coordinates system, determine the vehicle initial position road ID number, according to the spatial data (the path coordinate data that pretreated 5 meters sampling interval are discrete) and the attribute data of each node on the numbering of the start node of determining this road for initial road ID number, terminal node and this road; Calculate then the GPS locating information (vehicle is promptly located at that point for X, Y) nearest point to this road, and the judgment criterion that adopts when calculating is:
Wherein, (X Y) is the GPS locating information, (X
i, Y
i) be the coordinate points in the road.
Experiment shows, utilizes the data structure of road network topology relation to guarantee the real-time and the accuracy of information, obtained desirable locating effect, and map match can make being presented on the electronic chart of correct position of vehicle.The vehicle location of intersection is suitably carried out road matching to physical location skew collection point, for follow-up information utilization provides good basis also by the directional information in the synthetical collection information.
Step 4:
The dynamic traffic information collecting system is a kind of typical multisensor syste, and transport information has multi-source heterogeneous essential characteristic.Information fusion is one of core technology of traffic information acquisition system, by merging multi-source heterogeneous traffic data from multiple or a plurality of sensors, can obtain traffic flow condition information more accurately, thereby reduce the error in information processing, may occur, for the processing and utilization of transport information provides reliable basis.
Traditional traffic information fusion generally adopts the method for weighted sum, the fusion results of the traffic parameter measured value weighted sum that comes from each information acquisition equipment as this parameter.Such fusion method computing is simple, but the data precision after the uncertainty of its weights make to merge is low, in the basic demand that is difficult to satisfy the high-performance information acquisition system aspect the consistance of transport information and the integrality.Consider the good non-linear mapping capability of radial basis function neural network, and has simple in structure, a computing advantage fast, particularly have higher data precision and adaptive characteristic than conventional traffic information fusion method, can guarantee the consistance and the integrality of data message better, the present invention adopts based on the radial basis function neural network of the immune cluster basic tool as traffic information fusion, and corresponding Fusion Model is referred to as the immune cluster neural network.
The first step of using neural network realization traffic information fusion is structure and training network.Radial primary function network is a kind of three layers feedforward network, comprises input layer, output layer and hidden layer, and its structure as shown in Figure 5.
The node of input layer transmits input signal (measured value of each sensor traffic parameter promptly to be merged) to hidden layer, hidden layer radial basis function node carries out nonlinear transformation to input vector, its result obtains network output (being the fusion value of traffic parameter) in the output node weighted sum.
The hidden node of radial primary function network is made of radial function, and in this Fusion Model, its activation function adopts Gaussian function as follows:
X is the n dimension input vector of network in the formula, corresponding with traffic parameter to be merged (for example will use the magnitude of traffic flow that the above-mentioned network integration comes from three detecting devices, the input layer of network promptly has three neurons to constitute, at this moment n=3);
Be the output of i hidden node; C
iBeing the center of i hidden node basis function, is the column vector that has same dimension with X; σ
iBe the standard deviation of i node, represent the width of this basis function around central point; M is the number of hidden node.
If w
iBe the transmission weights of i hidden node to output node, then the output y of network can be represented by the formula:
For guaranteeing that network has desirable operating characteristic, need utilize the historical measurement data set pair network of traffic parameter to train, promptly for the center C of its basis function
iWith width cs
i, hidden layer is to the transmission weight w of output layer
iAnd the number m of hidden node is optimized.Because σ
iBe the parameter that can freely select, and after the hidden layer central point is determined, can calculate according to the number of hidden layer central point and hidden node (
c
Imax=Max (c
I1, c
I2..., c
In)), and output layer is to linear weight value w
iAdjustment can adopt the linear optimization method to obtain easily, therefore, the key of network training is reasonably to determine the center of basis function and the number of hidden node.
The general cluster of using training dataset in the hidden layer center of radial primary function network obtains, and is the training of the network hidden layer structure equivalence of cluster process to(for) training dataset promptly, and the cluster centre of training dataset is the center of hidden layer basis function at this moment.
The present invention is based on biological immune mechanism, provided a kind of improved artificial immunity clustering method.With other clustering methods according to the number (being the cluster numbers of data) of deviser's the artificial given hidden node of experience, that cluster obtains the center of basis function then is different, the present invention passes through the immune cluster to historical traffic data, simultaneously the hidden node number m and the basis function center C of optimization network
i, to obtain the having information fusion model of optimizing network structure.
The essence of immune cluster is based on the Immune Clone Selection and the immunological memory mechanism of Immune System, finds out one group of such data memory collection, and they are the one group of optimization antibody that has best affinity with antigen (corresponding data element to be clustered).In improved immune cluster algorithm, the affinity between definition antigen and the antibody be between the two distance against (under the one-dimensional case, being reciprocal relation), so the dimension of data memory collection is actually a kind of class number of optimizing cluster for data acquisition.Based on the training characteristic of radial primary function network, the cluster numbers of data set is the number of hidden nodes m of network.The center that not only can cluster goes out basis function, number that simultaneously can Automatic Optimal network hidden node, this is the maximum characteristics that immune cluster is different from other clustering method training radial primary function network.
In radially the immune cluster of base net network was trained, the training dataset of establishing network was the data acquisition of n * Q, and historical traffic data promptly to be clustered comes from n data detection resources, and each detection resources has Q data record.
For ease of describing, be defined as follows variable:
Ag: input antigen, historical traffic data set promptly to be clustered is expressed as form:
Ag
j＝[Ag
1j，Ag
2j，...，Ag
nj]
T∈R
n，j＝1，2，...，Q
Ab: initial antibodies, generally picked at random N row unit constitutes in Ag.Be expressed as form:
Ab＝[Ab
1，Ab
2，...，Ab
N]
Ab
i＝[Ab
1i，Ab
2i，...，Ab
ni]
T∈R
n，i＝1，2，...，N；N＜Q
E: the data memory collection, i.e. m network memory cell is expressed as form:
E＝[e
1，e
2，...，e
m]
e
k＝[e
1k，e
2k，...e
nk]
T∈R
n，k＝1，2，...，m；m＜＜Q
The number of cells m of data memory collection is the cluster numbers of training dataset behind the immune cluster, i.e. the hidden node number of radial primary function network.
Data
i: with memory antibody e
i(i=1,2 ..., m) have the antigen set of maximum affinity, be used to add up the ownership of the element of training dataset for cluster result, cluster finishes the back and calculates the center of gravity of each cluster data, the i.e. center of radial basis function by averaging method.
The immune cluster flow process of optimizing radial basis function hidden layer structure as shown in Figure 6, the concrete steps of algorithm are as follows:
N antibody A b of 4-1 random initializtion is as initial immunological network cell.
4-2 sets the cycle control parameter.
4-3 imports data to be clustered as antigen, to antigen Ag
j(j=1,2 ..., Q) carry out following operation:
4-3-1 computed range metric vector D
jWith affinity vector AF
jDistance between antigen and the antibody defines with Euclidean distance, that is:
According to the characteristics of cluster analysis, the affinity between antigen and the antibody adopts the following formula definition:
af
ij＝(1+d
ij)
-1
So: D
j=[d
1j, d
2j..., d
N, j]
T
AF
j＝[af
1j，af
2j，...，af
N，j]
T
4-3-2 selects t and Ag according to the ratio of optimum antibody selection rate opR from Ab
jAntibody with high-affinity is cloned increment, produces corresponding clone cell set C
Lj
By affinity series arrangement from small to large, the quantity of clone cell is calculated as follows to t selection cell:
In the formula: N
cClone cell sum for t antibody generation; α is a multiplier factor, is used to control the scale of clonal population; Int () is a bracket function.
4-3-3 uses following formula the antibody of cloning is carried out the random variation operation, realizes the maturation of affinity, produces to have the more antibody cell C of high-affinity
M:
C
M＝rand(C
R，N
R)
Rand (C in the formula
R, N
R) be random function, expression is from C
RIn randomly draw N
R(〉=1) individual variable; μ is the aberration rate of antibody, presses following formula and determines:
μ＝k×exp(-AF
v/γ)
In the formula, AF
v=AF
j/ || AF
j|| be the nominal value of parent affinity of antibody; K is a scale factor, and γ is the decay control coefrficient.The value principle of k and γ is individual within thresholding [0, the 1] scope of its permission for guaranteeing variation.
4-3-4 calculates C
MAntibody cell and antigen A g
jThe affinity vector
4-3-5 selects C according to selection ratio rsR once more
MIn several and antigen A g
jOptimization antibody with high-affinity is as part memory cell E
p
4-3-6 removes E
pMiddle similarity s
IjLess than immunological regulation threshold value σ
dAntibody, produce new memory collection E
k, realize that immune clone suppresses.
The similarity s of antibody
IjDescribe with the Euclidean distance between antibody, that is:
4-3-7 is part memory cell E
kMerge to memory cell set E (E ← [E; E
k]).
4-4 calculates the similarity vector S of each memory cell among the E, removes similarity s among the E
IjBe lower than immunosupress threshold value σ
sMemory cell, realize that the networks of different clone's collection suppress.
4-5 produces several antibody at random and replaces the lower individuality of affinity in the original antibody according to the poorest antibody selection rate wsR, embodies immune self organizing function.
The 4-6 variable is replaced, return 4-3, carry out follow-on e-learning, up to meeting the requirements of study algebraically, or satisfy the cluster iteration of setting and require (average affinity that has reached the number of predetermined memory cell or antigen and memory cell as iteration has reached the error range of being scheduled to etc.).
After said process finished, the output E of algorithm was the data memory collection [e of cluster data
1, e
2..., e
m], wherein the number m of memory cell is the cluster numbers of cluster data, the number of hidden nodes of corresponding radial primary function network.On this basis, obtain the cluster centre of data by following step.
4-7 input cluster data collection is as antigen, to antigen Ag
j(j=1,2 ..., Q) carry out following operation:
4-7-1 calculates antigen A g according to the definition of 4-3-1
jWith each memory antibody e
i(i=1,2 ..., affinity m);
4-7-2 is with Ag
jCharge to the memory cell e of affinity maximum with it
iPairing antigen set Data
i
The 4-8 gravity model appoach is asked the center of each data clusters, and formula is as follows:
N wherein
iBe data set Data
iThe number of middle antigen, the i.e. number of the cluster data that comprises in this generic.The center of data clusters is the central point c of radial basis function
i=[c
1i, c
2i... c
Ni]
T
More variable element is arranged, selection of parameter undue influence cluster effect in the above-mentioned algorithm.When the traffic data cluster was determined radially base net network hidden layer structure, value principle and the scope of each parameter are as follows: 1) initialization antibody number N rule of thumb selected, and size does not influence cluster result, generally gets N≤50; 2) for simplicity, the general value of clone multiplier factor α is 1; 3) optimum antibody selection rate opR generally gets 10%～20% of initial antibodies number; 4) the general value of the poorest antibody selection rate wsR should not surpass 10%; 5) selection rate rsR generally gets 10%～20% of clone cell number once more; 6) regulate threshold value σ
d=0.005～0.01, suppress threshold value σ
s=0.03～0.08.
Finish the structure and the hidden layer training of above-mentioned network, and the application training data set is to the output weight w of network
iAfter the linear optimization, the radial primary function network that obtains promptly can embed traffic information processing system as Fusion Model, realizes the fusion treatment to multi-source heterogeneous traffic parameter, and deposits traffic information database in.
Claims (3)
1. the method for acquiring dynamic traffic information based on middleware is characterized in that, its step is:
1) adopt serial interface communication mode and/or network communication mode to carry out the transmission of transport information;
2) the information acquisition port that utilizes the IDL (Interface Definition Language) IDL customization in the CORBA middleware Technology to be complementary with different transport information checkout equipments, identification and the data of standard from different checkout equipments realize the collection to the real-time dynamic information of road traffic flow, car speed, roadway occupancy;
3) all data that collect are carried out pre-service, and adopt the road matching algorithm of topological relation Network Based to carry out the map match of Floating Car;
The reparation and the missing value estimation that the pre-service of data are comprised abnormal data; Wherein, the data exception value adopts misdata to limit and handles, passing threshold method identification error data, promptly set corresponding data threshold according to data type, the threshold ratio up and down of flow, speed and occupation rate traffic parameter and its setting of checkout equipment collection, if measured value is then thought misdata not in the scope of last lower threshold value defined, it is repaired;
The threshold value of each traffic parameter is determined according to the planning index of the specified traffic capacity of road and the historical statistical data of each traffic parameter; Abnormal data is according to following formula reparation:
q
mod(t)＝αq
lw(t)+(1-α)q(t-1)
Wherein: q
Mod(t) be the abnormal data reparation value of traffic parameter q (t); q
Lw(t) be the measured value of last one all same working days, same period traffic parameter q (t); Q (t-1) was the measured value of a last period of traffic parameter q (t); The weights of α ∈ [0,1] for calculating, if certain traffic parameter has bigger time jitter characteristic, α gets less numerical value, otherwise α gets bigger numerical value;
The recognition methods of missing data is the data that the data definition that obtains become a certain moment in the certain hour section, then the time period of data is scanned, if do not obtaining data in the section sometime, thinks then that the data of this period have produced to lose;
The missing value estimation value is taken as the weighted sum of the measured value of a period on the measured value of all same working days, same period on the same traffic parameter and this traffic parameter, and computing formula is identical with the reparation of abnormal data;
4) utilize the immune cluster neural network, pretreated multi-source heterogeneous real-time dynamic traffic data are merged and deposit database in; Wherein, data fusion technology based on the immune cluster neural network, be the process of utilizing radial basis function neural network that multi-source heterogeneous real-time dynamic traffic data are merged, wherein the hidden layer structure of radial basis function neural network is determined by the data clustering method based on the artificial immunity theory.
2. the method for acquiring dynamic traffic information based on middleware as claimed in claim 1, it is characterized in that, described step 2) utilize the IDL (Interface Definition Language) IDL in the CORBA middleware Technology to carry out unified organizational system in from the data layout of different transport information checkout equipments, collection port and each transport information checkout equipment are complementary, gather real-time dynamic information from road traffic flow, car speed and the roadway occupancy of different transport information checkout equipments;
Wherein, to doing as giving a definition from the data layout of different transport information checkout equipments: the definition event structure comprises incident sequence number and event description; The checkout equipment type comprises coil checker, microwave detector, video detector and Floating Car checkout equipment; The external communication protocol that checkout equipment adopts comprises RS-232/422/485,802.1, GPRS and Ethernet; The essential information of checkout equipment comprises: brand, manufacturer, unit type, device type, device numbering and communication protocol; Discharge pattern comprises forward flow and reverse flow; Speed type comprises forward direction speed and inverted speed; The occupation rate type comprises forward occupation rate and reverse occupation rate; The checkout equipment state comprise success, invalid, finish and do not finish; The checkout equipment acquisition parameter comprises flow information, velocity information and occupation rate information; The checkout equipment operation comprises job title and job description;
To the following operation of information acquisition port definition: the checkout equipment initialization, and be connected with collection terminal; Judge whether checkout equipment is in active state; Disconnect and connecting; Suspend; Get back to initial bit; Upload operation; Download operation; Gather the road traffic flow; Collection vehicle speed; Gather roadway occupancy.
3. the method for acquiring dynamic traffic information based on middleware as claimed in claim 1, it is characterized in that, in the described step 3), map-matching method is: the vehicle location information that GPS is received is planned for map coordinates system, determine vehicle initial position road ID number, according to initial road ID number, determine the spatial data and the attribute data of each node on start node numbering, terminal node numbering and this road of this road; Calculate the GPS locating information (vehicle is promptly located at that point for X, Y) nearest point to this road, and the judgment criterion that adopts when calculating is:
Wherein (X Y) is GPS locating information, (X
i, Y
i) be the coordinate points in the road.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2009102565373A CN101719315B (en) | 2009-12-23 | 2009-12-23 | Method for acquiring dynamic traffic information based on middleware |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2009102565373A CN101719315B (en) | 2009-12-23 | 2009-12-23 | Method for acquiring dynamic traffic information based on middleware |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101719315A CN101719315A (en) | 2010-06-02 |
| CN101719315B true CN101719315B (en) | 2011-06-01 |
Family
ID=42433883
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2009102565373A Expired - Fee Related CN101719315B (en) | 2009-12-23 | 2009-12-23 | Method for acquiring dynamic traffic information based on middleware |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101719315B (en) |
Families Citing this family (41)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101968927B (en) * | 2010-09-21 | 2015-07-01 | 隋亚刚 | Detection system for acquiring traffic flow by using high-definition video and microwave detection selectively |
| CN102254431B (en) * | 2011-06-14 | 2014-05-21 | 上海雷腾软件有限公司 | Vehicle position data acquiring and processing method and system |
| CN102355471B (en) * | 2011-06-14 | 2013-10-23 | 上海雷腾软件有限公司 | Dual-source distributed method for collection of car position data and system thereof |
| CN102288183B (en) * | 2011-06-22 | 2013-04-24 | 北京农业信息技术研究中心 | System and method for positioning address for reporting rural event |
| CN102521981A (en) * | 2011-12-20 | 2012-06-27 | 青岛海信网络科技股份有限公司 | Computation method for traffic situation based on information-oriented middleware |
| CN102881125B (en) * | 2012-09-25 | 2014-06-18 | 杭州立高科技有限公司 | Alarm monitoring system based on multi-information fusion centralized processing platform |
| CN103036874B (en) * | 2012-11-28 | 2015-10-28 | 大连理工大学 | The guard method of prevention data injection attacks in gathering for car networking data |
| CN103593973B (en) * | 2012-12-18 | 2016-07-06 | 北京科技大学 | A kind of urban road traffic situation assessment system |
| CN103049522B (en) * | 2012-12-19 | 2016-01-20 | 中国电力科学研究院 | A kind of method automatically substituted for electricity multi-data source |
| CN103413428A (en) * | 2013-06-27 | 2013-11-27 | 北京交通大学 | Expression method of road traffic information credibility space characteristics based on sensor network |
| CN103346932B (en) * | 2013-07-10 | 2018-04-24 | 中兴智能交通股份有限公司 | Method and device for judging state of traffic equipment based on data audit |
| CN104424799B (en) * | 2013-09-05 | 2017-07-07 | 吴汉才 | A kind of real-time dynamic road condition collection and the system and method for service |
| CN103730016B (en) * | 2013-12-17 | 2017-02-01 | 深圳先进技术研究院 | Traffic information publishing system |
| DE102014208524A1 (en) * | 2014-05-07 | 2015-11-12 | Robert Bosch Gmbh | LOCAL TRANSPORTATION ANALYSIS WITH DETECTION OF A TRAFFIC PATH |
| CN104778834B (en) * | 2015-01-23 | 2017-02-22 | 哈尔滨工业大学 | Urban road traffic jam judging method based on vehicle GPS data |
| CN105118289A (en) * | 2015-06-17 | 2015-12-02 | 河南理工大学 | Traffic situation assessment method |
| CN105225486A (en) * | 2015-10-09 | 2016-01-06 | 哈尔滨工业大学深圳研究生院 | Fill up the method and system of disappearance floating car data |
| CN105702041A (en) * | 2016-04-21 | 2016-06-22 | 东南大学 | Highway multisource data fusion state estimation system based on neural network and method thereof |
| CN106446091A (en) * | 2016-09-13 | 2017-02-22 | 北京协力筑成金融信息服务股份有限公司 | Preprocessing method and device for multi-source time series data |
| CN106530693B (en) * | 2016-11-02 | 2019-02-05 | 浙江大学 | Urban expressway traffic state based on fused data rolls estimation method |
| JP6988890B2 (en) * | 2017-05-25 | 2022-01-05 | 日本電気株式会社 | Processing equipment, processing methods and programs |
| CN107633692A (en) * | 2017-09-29 | 2018-01-26 | 河南理工大学 | A kind of city expressway Entrance ramp MFA control method |
| CN109842649A (en) * | 2017-11-27 | 2019-06-04 | 北京通号国铁城市轨道技术有限公司 | A kind of urban rail transit equipment health control method and device |
| CN108182817A (en) * | 2018-01-11 | 2018-06-19 | 北京图森未来科技有限公司 | Automatic Pilot auxiliary system, trackside end auxiliary system and vehicle-mounted end auxiliary system |
| CN108550262B (en) * | 2018-06-01 | 2021-06-11 | 中物汽车电子扬州有限公司 | Urban traffic sensing system based on millimeter wave radar |
| CN108831168B (en) * | 2018-06-01 | 2020-10-02 | 武汉泰坦智慧科技有限公司 | Traffic signal lamp control method and system based on visual identification of associated intersection |
| CN109063752B (en) * | 2018-07-17 | 2022-06-17 | 华北水利水电大学 | Multi-source high-dimensional multi-scale real-time data stream sorting method based on neural network |
| CN109272746B (en) * | 2018-08-20 | 2021-06-08 | 广东交通职业技术学院 | MFD estimation method based on BP neural network data fusion |
| CN109213755B (en) * | 2018-09-30 | 2019-09-24 | 长安大学 | A kind of traffic flow data cleaning and restorative procedure based on Time-space serial |
| US20200133267A1 (en) * | 2018-10-25 | 2020-04-30 | GM Global Technology Operations LLC | Middleware support for fault-tolerant execution in an adaptive platform for a vehicle |
| CN109544901A (en) * | 2018-11-26 | 2019-03-29 | 南京行者易智能交通科技有限公司 | A kind of Research on Intelligent Scheduling of Public Traffic Vehicles method and device based on history passenger flow big data |
| CN109933624A (en) * | 2018-12-28 | 2019-06-25 | 曙光信息产业(北京)有限公司 | Aviation emergency management and rescue data collection system and method |
| CN110096497A (en) * | 2019-03-28 | 2019-08-06 | 中国农业科学院农业信息研究所 | A kind of agricultural output data intelligence cleaning method and system |
| CN110164127B (en) * | 2019-04-04 | 2021-06-25 | 中兴飞流信息科技有限公司 | Traffic flow prediction method and device and server |
| CN110647518B (en) * | 2019-11-28 | 2020-04-28 | 成都四方伟业软件股份有限公司 | Data source fusion calculation method, component and device |
| CN111063091B (en) * | 2019-12-27 | 2023-01-20 | 惠州视维新技术有限公司 | Play yard management method, system, terminal equipment and storage medium |
| CN111598210B (en) * | 2020-04-30 | 2023-06-02 | 浙江工业大学 | Anti-attack defense method for anti-attack based on artificial immune algorithm |
| CN112270840B (en) * | 2020-10-23 | 2021-09-10 | 清华大学 | Information credible identification method based on single-vehicle motion characteristics under vehicle-road cooperative environment |
| CN113259900B (en) * | 2021-05-27 | 2021-10-15 | 华砺智行(武汉)科技有限公司 | Distributed multi-source heterogeneous traffic data fusion method and device |
| CN113869571B (en) * | 2021-09-22 | 2024-04-26 | 国网天津市电力公司电力科学研究院 | Intelligent energy building comprehensive information physical fusion method based on heterogeneous data |
| CN114333301B (en) * | 2021-12-06 | 2023-09-26 | 北京东土正创科技有限公司 | Traffic signal control optimization method, system and traffic signal optimization equipment |
-
2009
- 2009-12-23 CN CN2009102565373A patent/CN101719315B/en not_active Expired - Fee Related
Also Published As
| Publication number | Publication date |
|---|---|
| CN101719315A (en) | 2010-06-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101719315B (en) | Method for acquiring dynamic traffic information based on middleware | |
| CN102163225B (en) | A fusion evaluation method of traffic information collected based on micro blogs | |
| CN110008413B (en) | Traffic travel problem query method and device | |
| Kong et al. | Big data‐driven machine learning‐enabled traffic flow prediction | |
| CN106781509B (en) | A kind of collaborative urban road congestion detection method based on V2V | |
| CN107463940A (en) | Vehicle type recognition method and apparatus based on data in mobile phone | |
| CN108513676A (en) | A kind of road conditions recognition methods, device and equipment | |
| Rabie et al. | A new outlier rejection methodology for supporting load forecasting in smart grids based on big data | |
| CN113591380A (en) | Traffic flow prediction method, medium and equipment based on graph Gaussian process | |
| CN111310294A (en) | Method for establishing and issuing evaluation index system of traffic management and control service index | |
| CN110598917B (en) | Destination prediction method, system and storage medium based on path track | |
| Tu et al. | Research on intelligent calculation method of intelligent traffic flow index based on big data mining | |
| CN111899511A (en) | Bus arrival time prediction method for AVL data of collinear line | |
| Zou et al. | The method of dynamic identification of the maximum speed limit of expressway based on electronic toll collection data | |
| Seo et al. | LoRa based architecture for smart town traffic management system | |
| Hui et al. | Hotspots identification and classification of dockless bicycle sharing service under electric fence circumstances | |
| Lin et al. | Insights into Travel Pattern Analysis and Demand Prediction: A Data-Driven Approach in Bike-Sharing Systems | |
| US20230153742A1 (en) | Methods for shared bicycle delivery and operation area planning in smart cities and internet of things (iot) systems thereof | |
| Ding et al. | A deep learning based traffic state estimation method for mixed traffic flow environment | |
| Li | Research on scheduling algorithm of agricultural machinery cooperative operation based on particle swarm neural network | |
| Guo et al. | Research on optimization model of multisource traffic information collection combination based on genetic algorithm | |
| Lan et al. | Inferring alighting bus stops from smart card data combined with cellular signaling data | |
| CN110444038A (en) | Bus dispatching method based on big data | |
| Jiang | [Retracted] Highway Traffic Flow Prediction Model Construction Based on the Gray Theory and BP Neural Network | |
| CN117708260B (en) | Smart city data linkage updating method and system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C17 | Cessation of patent right | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20110601 Termination date: 20121223 |