CN101739824B - Data fusion technology-based traffic condition estimation method - Google Patents

Data fusion technology-based traffic condition estimation method Download PDF

Info

Publication number
CN101739824B
CN101739824B CN2009101986924A CN200910198692A CN101739824B CN 101739824 B CN101739824 B CN 101739824B CN 2009101986924 A CN2009101986924 A CN 2009101986924A CN 200910198692 A CN200910198692 A CN 200910198692A CN 101739824 B CN101739824 B CN 101739824B
Authority
CN
China
Prior art keywords
data
traffic
average velocity
traffic condition
road
Prior art date
Application number
CN2009101986924A
Other languages
Chinese (zh)
Other versions
CN101739824A (en
Inventor
陈林
魏淑桃
石林祥
Original Assignee
上海第二工业大学
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 上海第二工业大学 filed Critical 上海第二工业大学
Priority to CN2009101986924A priority Critical patent/CN101739824B/en
Publication of CN101739824A publication Critical patent/CN101739824A/en
Application granted granted Critical
Publication of CN101739824B publication Critical patent/CN101739824B/en

Links

Abstract

The invention discloses a data fusion technology-based traffic condition estimation method. The method comprises the following steps: acquiring current traffic condition information by using a GPS, an annular vehicle detector and other sensors based on the principle that much information is helpful for estimating the traffic condition more accurately, then preprocessing the traffic condition information, and performing fusion processing on the traffic condition information and sensor data of weather, time period and different moments together by adopting a Bayesian network to acquire accurate information of the current traffic condition. Besides the current sensor date, the method takes multiple other factors influencing the traffic condition into consideration, and performs fusion processing so as to avoid inaccurate judgment problem brought by single sensor information. The traffic condition estimation method is particularly suitable for traffic environments of large cities.

Description

Traffic condition estimation method based on the data fusion technology
Technical field
The present invention relates to a kind of traffic condition estimation method, relate to particularly that to overcome the state estimation that causes in the individual data source not accurate enough, a kind of traffic condition estimation method of big these shortcomings of error based on the data fusion technology.
Background technology
The development of world car industry and family-sized car universal day by day, the problem of blocking up of road traffic becomes the major issue of urban transportation; The progress at full speed of infotech is that comprehensive transport solution problem has been brought opportunity on the other hand.Be exactly under this background; Advanced traffic information management system has received concern widely prior to the other system of intelligent transportation system; All obtained development at full speed in countries in the world, be applied to dynamic route planning, dynamic navigation, road network and coordinate various aspects such as traffic signal system, dynamic traffic scheduling.Wherein, the estimation of road net traffic state is the key components of traffic information management system in real time.
Road net traffic state is estimated in real time relevant with the transport information that is adopted, different transport information has determined estimation approach and precision.At present; Many correlative studys have been arranged in the world, and wherein, that representative is Martin L.Hazclton (" Estimating Vehicle Speed from Count andOccupancy data; Journal of Data Science 2 (2004), 231-244 ").MartinL.Hazclton considers effectively and modeling to have handled Road Detection loop data error rate big; The problem that reliability is low; And the result is gratifying, but he is the research of on the expressway, carrying out, and only is applicable to that traffic flow is the situation of continuous stream; And the traffic flow in city stops between being, the traffic flow modes that is not suitable for city road network is estimated.Utilize the detection ring data that city road network is carried out traffic flow and estimate urban infrastructure is had relatively high expectations, often get less than enough desired datas in a lot of cities, and the high problem of error rate can not get solving effectively.
In patent documentation " method for estimating city road network traffic flow state, application number: 200510026214.7 ", the applicant has proposed a kind of method for estimating state of the traffic control information based on gps data and self-adaptation traffic control system (SCATS).But use gps data merely, have a lot of error situations, cause erroneous judgement easily.Such as; Gps data shows that Vehicle Speed is 0; System generally is judged as traffic congestion, but current vehicle also possibly be because on-board and off-board, vehicle GPS data that perhaps cause because of red light and show that Vehicle Speed is 0; Be not real traffic congestion, the misjudgment of the easy like this traffic behavior that causes.And the information of traffic control system and be inconvenient to obtain.
Summary of the invention
To the problems referred to above; Fundamental purpose of the present invention is the deficiency to prior art; Propose a kind of new traffic flow modes method of estimation that is used for city road network based on data fusion, it is not accurate enough to overcome the state estimation that causes in the individual data source, the shortcoming that error is bigger.
The present invention solves above-mentioned technical matters through following technical proposals: in order to realize such purpose; The data source of technical scheme of the present invention comes from; One type from sensing data, comprise the GPS receiver that loads on annular wagon detector and the vehicle.The annular wagon detector can obtain the mean velocity information that current road vehicle goes; The GPS receiver that loads on the vehicle also can obtain the car speed information of current travels down, if receive the GPS information that many cars on the same highway section send over simultaneously, just can obtain the mean velocity information of vehicle on this highway section.
Another kind of information is and the closely-related non-sensor information of traffic, comprises the current weather situation, current information of time.These information are also to the very important influence of whether having blocked up of a road.
Last category information is the sensor historical information, comprises the vehicle mean velocity information of the past GPS receiver acquisition in the short time now and the road vehicle mean velocity information that annular wagon detector obtains.
This method mainly comprises following step:
(1). through the traffic related information on sensor acquisition road surface;
(2). the sensing data to sensor acquisition in the step (1) carries out pre-service;
(3). the sensing data of gathering and current weather, time are adopted Bayesian network method, data are carried out fusion treatment;
(4). through the data estimation traffic that obtains after the fusion treatment.
Describe respectively below:
1. gps data is carried out pre-service:
Be regarded as the current real-time GPS data that can provide the vehicle GPS receiver that comprises distance, time, velocity information to pass over the sampled point of wagon flow on each oriented highway section.The pre-service of gps data mainly comprises at first and to carry out road matching according to the latitude and longitude information in the gps data, confirms that gps data comes from the vehicle on that highway section; To try to achieve the average velocity of this highway section vehicle according to the car speed information in all gps datas on this highway section then.
2. annular wagon detector data are carried out pre-service:
Because annular wagon detector can directly provide the average velocity of the vehicle ' on the highway section, so no longer need other processing.
3. delimit the congestion in road state by average velocity:
With t 0The average velocity of each oriented highway section road direction is the delimitation that index is carried out the congestion in road state in the city road network constantly.The average velocity in highway section in the city road network is divided into five speed class, corresponding respectively unimpeded, more unimpeded, obstructed, five kinds of congestion in road states block up, seriously block up.Judge the congestion status in each oriented highway section according to the corresponding residing speed class of average velocity in each oriented highway section.
4. the Bayesian network that is used for data fusion makes up:
According to the cause-effect relationship between the data, make up Bayesian network.Each node in the Bayesian network is represented a stochastic variable, and arrow points to the result by reason.The probability distribution table of the stochastic variable of each node representative then obtains through observing statistics and experiment.
5. based on the data fusion of variable method of elimination
It is a kind of Bayesian network inference method commonly used that variable disappears that unit sends out.Progressively cancellation does not belong to query interface, and just its dependent variable of traffic variable finally obtains on the basis of known evidence variable, and query interface is the probability distribution situation of traffic just.Traffic after estimating to obtain merging.
Positive progressive effect of the present invention is: the traffic condition estimation method based on the data fusion technology provided by the invention has following advantage:
1. this method has also been considered a plurality of other factors that influence traffic, and has been carried out fusion treatment except current sensor data, and it is not accurate enough to have overcome the state estimation that causes in the individual data source, the shortcoming that error is bigger.
Description of drawings
Fig. 1 is the FB(flow block) of urban road network traffic situation method of estimation;
Fig. 2 is the Bayesian network synoptic diagram that is used for data fusion.
Embodiment
Provide preferred embodiment of the present invention below in conjunction with accompanying drawing, to specify technical scheme of the present invention.The FB(flow block) of Fig. 1 urban road network traffic situation method of estimation, as shown in Figure 1, the traffic condition estimation method based on the data fusion technology provided by the invention comprises following a few step:
(1). through the traffic related information on sensor acquisition road surface;
(2). the sensing data to sensor acquisition in the step (1) carries out pre-service;
(3). the sensing data of gathering and current weather, time are adopted Bayesian network method, data are carried out fusion treatment;
(4). through the data estimation traffic that obtains after the fusion treatment.
Describe respectively below:
1. data pre-service
Input data of the presently claimed invention are as shown in Figure 2; Comprise that GPS measures average velocity, annular wagon detector measuring vehicle average velocity and a last moment annular wagon detector measuring vehicle average velocity of current average velocity, a last GPS measurement constantly, and weather and time period data.Average velocity v can be divided into 5 speed class according to the actual traffic state; Traffic behavior seriously blocks up corresponding to v≤10 kilometer/hour, blocks up corresponding to 10<v≤20, and is not smooth corresponding to 20<v≤30; More unimpeded corresponding to 30<v≤50, unobstructed corresponding to v>50.Weather data is divided into abominable, relatively poor, general and better according to the influence degree to traffic.Time period is divided into peak period, crowded period, general period and unimpeded period according to the traffic impact degree.
2. Bayesian network makes up
The bayesian network structure that makes up is as shown in Figure 2, and arrow is represented cause-effect relationship to be arranged between the two if raising data fusion precision can increase more data; Such as increasing more historical sensing data; Perhaps increase more sensing data type, for example, video data or the like.For confirming of the probability distribution of each node in the Bayesian network, the method that adopts actual measurement and expert to authorize.Through the observation long-term to a certain highway section, can confirm, under different weather conditions, the situation of traffic, thus can confirm the conditional probability distribution from weather to the traffic.
3. based on the data fusion of variable method of elimination
The input fusion process is following:
1. initialization
Construct an initial intermediate function formation, the corresponding function of all nodes in the Bayesian network is put into wherein; Variablees such as the sensing data that then initial function formation basis is obtained, weather are set its initial value; Construct the metavariable sequence of waiting to disappear then, this sequence does not comprise known preset value variable and traffic variable to be asked.
The metavariable that is still waiting to disappear in the metavariable sequence if 2. wait to disappear then repeat this step
Construct a new function formation, the function relevant with the metavariable of waiting to disappear in original function formation all moved into new function formation; Construct a new function, this function is the result of calculation of new function formation; To all involved in new function formation variablees assignment successively, promptly from (0,0 ... 0) always assignment to (m, n ... Z) (m wherein, n ....; Z etc. are the maximum occurrences classification numbers of each variable), each assignment is calculated company's value of taking advantage of of new function formation, adding up afterwards by the value of the metavariable of waiting to disappear this companys value of taking advantage of, the value of gained is exactly the functional value of new function under current assignment condition; The new function that calculating is finished joins in original function formation;
3. all functions in the present function formation are merged into a function and carry out the normalization processing, last resulting function is exactly a probability distribution of under current known conditions, waiting to ask traffic, and the traffic that promptly obtains current maximum probability is estimated.
More than show and described ultimate principle of the present invention and principal character and advantage of the present invention.The technician of the industry should understand; The present invention is not restricted to the described embodiments; That describes in the foregoing description and the instructions just explains principle of the present invention, and under the prerequisite that does not break away from spirit and scope of the invention, the present invention also has various changes and modifications; These variations and improvement all fall in the scope of the invention that requires protection, and the present invention requires protection domain to be defined by appending claims and equivalent thereof.

Claims (4)

1. traffic condition estimation method based on data fusion technology, it is characterized in that: said method comprises the steps:
(1). through the traffic related information on sensor acquisition road surface;
(2). the sensing data to sensor acquisition in the step (1) carries out pre-service;
A. gps data is carried out pre-service:
At first, carry out road matching, confirm gps data comes from the vehicle on which bar highway section according to the latitude and longitude information in the gps data; To try to achieve the average velocity of this highway section vehicle according to the car speed information in all gps datas on this highway section then;
B. annular wagon detector data are carried out pre-service:
Because annular wagon detector can directly provide the average velocity of the vehicle ' on the highway section, so no longer need other processing;
C. delimit the congestion in road state by average velocity:
With t 0The average velocity of each oriented highway section road direction is the delimitation that index is carried out the congestion in road state in the city road network constantly; The average velocity in highway section in the city road network is divided into five speed class; Corresponding respectively unimpeded, more unimpeded, obstructed, the five kinds of congestion in road states that block up, seriously block up are judged the congestion status in each oriented highway section according to the corresponding residing speed class of average velocity in each oriented highway section;
(3). the sensing data of gathering and current weather, time are adopted Bayesian network method, data are carried out fusion treatment;
According to the cause-effect relationship between the data; Make up Bayesian network; Each node in the Bayesian network is represented a stochastic variable, and arrow points to the result by reason, and the probability distribution table of the stochastic variable of each node representative then obtains through observing statistics and experiment;
(4). through the data estimation traffic that obtains after the fusion treatment;
1. initialization
Construct an initial intermediate function formation, the corresponding function of all nodes in the Bayesian network is put into wherein; The sensing data, the weather variable that then initial function formation basis are obtained are set its initial value; Construct the metavariable sequence of waiting to disappear then, this sequence does not comprise known preset value variable and traffic variable to be asked;
The metavariable that is still waiting to disappear in the metavariable sequence if 2. wait to disappear then repeat this step
Construct a new function formation, the function relevant with the metavariable of waiting to disappear in original function formation all moved into new function formation; Construct a new function, this function is the result of calculation of new function formation; To all involved in new function formation variablees assignment successively, promptly from 0,0 ... 0 always assignment to m, n ..., z; M wherein, n ... Z is the maximum occurrences classification number of each variable, and each assignment is calculated company's value of taking advantage of of new function formation, and adding up afterwards by the value of the metavariable of waiting to disappear this companys value of taking advantage of, the value of gained is exactly the functional value of new function under current assignment condition; The new function that calculating is finished joins in original function formation;
3. all functions in the present function formation are merged into a function and carry out the normalization processing, last resulting function is exactly a probability distribution of under current known conditions, waiting to ask traffic, and the traffic that promptly obtains current maximum probability is estimated.
2. the traffic condition estimation method based on the data fusion technology according to claim 1; It is characterized in that: said sensor comprises two types: one type of GPS receiver that is on the automobile to be installed, another kind of is to be embedded in the underground annular wagon detector of road.
3. the traffic condition estimation method based on data fusion technology according to claim 1 is characterized in that: said sensing data had both comprised the sensing data of current time also comprising the sensor historical data.
4. the traffic condition estimation method based on the data fusion technology according to claim 2; It is characterized in that: said data of carrying out fusion treatment comprise that GPS measures current average velocity, last average velocity, annular wagon detector measuring vehicle average velocity and the last moment annular wagon detector measuring vehicle average velocity measured of GPS constantly, and weather and time period data.
CN2009101986924A 2009-11-12 2009-11-12 Data fusion technology-based traffic condition estimation method CN101739824B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN2009101986924A CN101739824B (en) 2009-11-12 2009-11-12 Data fusion technology-based traffic condition estimation method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN2009101986924A CN101739824B (en) 2009-11-12 2009-11-12 Data fusion technology-based traffic condition estimation method

Publications (2)

Publication Number Publication Date
CN101739824A CN101739824A (en) 2010-06-16
CN101739824B true CN101739824B (en) 2012-11-14

Family

ID=42463253

Family Applications (1)

Application Number Title Priority Date Filing Date
CN2009101986924A CN101739824B (en) 2009-11-12 2009-11-12 Data fusion technology-based traffic condition estimation method

Country Status (1)

Country Link
CN (1) CN101739824B (en)

Families Citing this family (17)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102646332B (en) * 2011-02-21 2014-03-12 日电(中国)有限公司 Traffic state estimation device and method based on data fusion
CN102289932B (en) * 2011-06-17 2013-10-23 同济大学 Dynamic OD (Origin Destination) matrix estimating method based on AVI (Automatic Vehicle Identification) device
CN102436740A (en) * 2011-09-29 2012-05-02 东南大学 Automatic detection method of traffic incident on highway
CN102932812B (en) * 2012-11-06 2014-11-19 武汉大学 Vehicle sensor concurrent monitoring method facing road conditions
CN103903436A (en) * 2012-12-28 2014-07-02 上海优途信息科技有限公司 Expressway traffic jam detecting method and system based on floating car
CN104933856B (en) * 2014-03-17 2017-08-25 径卫视觉科技(上海)有限公司 Road conditions real-time evaluation system and method
DE102014205180A1 (en) * 2014-03-20 2015-09-24 Robert Bosch Gmbh Method and device for operating a vehicle
CN103839412B (en) * 2014-03-27 2015-12-02 北京建筑大学 A kind of crossing dynamic steering ratio combination method of estimation based on Bayes's weighting
CN105096590B (en) * 2014-04-23 2019-07-26 株式会社日立制作所 Traffic information creating method and traffic information generating device
CN104021672B (en) * 2014-06-09 2016-09-07 北京世纪高通科技有限公司 A kind of method and apparatus obtaining traffic congestion index
CN105808918B (en) * 2015-08-16 2019-04-05 北京数行健科技有限公司 A kind of road full flow Backstipping design based on data fusion
CN106251630B (en) * 2016-10-13 2018-09-07 东南大学 A kind of progressive Extended Kalman filter traffic status of express way method of estimation based on multi-source data
CN108275158B (en) * 2017-01-05 2020-09-15 大唐高鸿信息通信(义乌)有限公司 Driving behavior evaluation method of vehicle-mounted short-distance communication network
CN108734955A (en) * 2017-04-14 2018-11-02 腾讯科技(深圳)有限公司 The method and device of predicting road conditions state
CN108109382B (en) * 2018-02-05 2020-08-04 青岛大学 Congestion point, congestion line and congestion area discovery method based on composite network
CN111326002A (en) * 2020-02-26 2020-06-23 公安部交通管理科学研究所 Prediction method, device and system for environment perception of automatic driving automobile
CN111429728A (en) * 2020-05-06 2020-07-17 辽宁百思特达半导体科技有限公司 Intelligent traffic system based on intelligent lamp pole and method thereof

Also Published As

Publication number Publication date
CN101739824A (en) 2010-06-16

Similar Documents

Publication Publication Date Title
CN105374211B (en) The system and method for driving risk and aiding in vehicle insurance price is calculated based on multi-source data
CN107000750B (en) Apparatus and method for vehicle economy improvement
Seo et al. Estimation of flow and density using probe vehicles with spacing measurement equipment
CN103500503B (en) A kind of accurate road condition analyzing method and system based on mass-rent pattern
CN106297288B (en) A kind of acquisition of bus passenger passenger flow data and analysis method
Cortés et al. Commercial bus speed diagnosis based on GPS-monitored data
Haghani et al. Data collection of freeway travel time ground truth with bluetooth sensors
WO2015051718A1 (en) Dynamic track navigation method and cloud platform
CN104809879B (en) Expressway road traffic state estimation method based on dynamic Bayesian network
CN102968901B (en) Method for acquiring regional congestion information and regional congestion analyzing device
US8694241B1 (en) Visualization of traffic patterns using GPS data
JP4716115B2 (en) Traffic flow parameter calculation system, method and program
JP5374067B2 (en) Traffic condition simulation apparatus and program
CN103258430B (en) Road traveling time calculating and traffic road condition judging method and road traveling time calculating and traffic road condition judging device
Marfia et al. Vehicular congestion detection and short-term forecasting: a new model with results
CN104809878B (en) Method for detecting abnormal condition of urban road traffic by utilizing GPS (Global Positioning System) data of public buses
CN104933856B (en) Road conditions real-time evaluation system and method
CN102419905B (en) Traffic-wave theory-based traffic influence area determining method of expressway accidents
CN109410586B (en) Traffic state detection method based on multi-metadata fusion
WO2019001044A1 (en) Mobile phone data-based vehicle model identification method and device
CN104376729B (en) A kind of highway based on truck traffic is to reminding method congested to track
JP5424754B2 (en) Link travel time calculation device and program
CN105185145A (en) Intelligent dynamic road condition map navigation system and method
Li et al. Link travel time estimation using single GPS equipped probe vehicle
CN104590275B (en) A kind of driving behavior analysis method

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
CF01 Termination of patent right due to non-payment of annual fee
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20121114

Termination date: 20151112