CN103646534B - A kind of road real-time traffic accident risk control method - Google Patents
A kind of road real-time traffic accident risk control method Download PDFInfo
- Publication number
- CN103646534B CN103646534B CN201310596435.2A CN201310596435A CN103646534B CN 103646534 B CN103646534 B CN 103646534B CN 201310596435 A CN201310596435 A CN 201310596435A CN 103646534 B CN103646534 B CN 103646534B
- Authority
- CN
- China
- Prior art keywords
- traffic
- road
- real
- data
- classification
- 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
Landscapes
- Traffic Control Systems (AREA)
Abstract
The invention discloses a kind of road real-time traffic accident risk control method, is a kind of road real-time traffic accident risk prediction based on multi-class support vector machine and control method, can be used to the possibility that traffic hazard occurs in predicted detection section.To detecting section foundation based on the accident prediction model of multi-class support vector machine, bringing the real-time traffic characteristic parameter of collection into accident prediction model, judging whether the risk that traffic hazard occurs.This method utilizes the real-time traffic characteristic parameter of collection to predict contingent traffic hazard, and has good precision of prediction, overcomes prior art and utilizes collection meter statistic to analyze the technological deficiency of traffic safety existence with not enough.This method is judging that the risk that all kinds of traffic hazard occurs has actual engineer applied value.
Description
Technical field
The invention belongs to intelligent traffic administration system and control technology field, use support vector machine to existing traffic hazard learning classification model and then carry out accident risk discrimination and classification, propose a kind of method of discrimination to certain section road traffic accident occurrence risk and control solution.
Background technology
The increase of automobile pollution and day by day frequently road traffic trip and logistics transportation, bring about a prosperous economy and in life while, day by day serious pressure is brought to traffic safety environment, objectively also active demand is proposed to the research and development of the safe and smooth without hindrance correlation technique of road traffic environment.
Along with the lifting of traffic safety management level and the development of vehicle active safety technologies, transportation security environments is greatly improved; Especially high speed road traffic environment, according to statistics national freeway traffic environment, national traffic accidents quantity is since two thousand three in progressively downward trend according to statistics, and accumulative fall reaches 70%.But downtrending is in recent years gradually steady, and the highway major motor vehicle security incident number of casualties maintains a high position always.Current traffic accidents form main manifestations is: rear-end impact, hit fixture and stationary vehicle, overturning, side impact, just brush up against scraping.
Summary of the invention
The problem to be solved in the present invention is: under transport information Real time data acquisition and storage have become this background relatively easy, set up the relation between real time traffic data feature and street accidents risks, propose the road real-time traffic accident risk prediction based on multi-class support vector machine and control method.
Technical scheme of the present invention is: based on multi-class support vector machine road real-time traffic accident risk prediction control method, road real-time traffic accident risk prediction and Controlling model are set up to surveyed area, Real-time Road accident risk identification and classification model is substituted into after real time traffic data information is proposed feature, whether have an accident and accident pattern according to Real-time Road accident risk identification and classification model Output rusults predicted detection region, propose corresponding control measure thus.From two aspects expansion below: (1), street accidents risks affair character extract; (2), road traffic risk profile and control.
Technical scheme: for solving the problems of the technologies described above, the present invention adopts based on multi-class support vector machine road real-time traffic accident risk prediction and control method, comprises the following steps:
Steps A) gather under normal traffic states and before having accident to occur t second (0.01≤t≤0.1) state under data sample N group { X
1, X
2..., X
n.Each X
i(I=1,2,3 ..., N) comprise people, car, road, environmental information: according to time of origin and the generation section of traffic hazard, there is the sex x of people from section in random acquisition traffic hazard
1, driving efficiency x
2; The spacing x of car
3, speed of a motor vehicle x
4, vehicle condition x
5; The type x of road
6, pavement behavior x
7, traffic x
8; And the weather x on the same day
9, pedestrian situation x
10, visibility x
11.
Step B) gather and above-mentioned data sample { X
1, X
2..., X
ncorresponding traffic hazard classification sample { Y
1, Y
2..., Y
n, each Y
i(I=1,2,3 ..., N) correspond to { w
1, w
2, w
3in a certain item.Wherein w
1represent rear-end impact accident, w
2represent and hit fixture and stationary vehicle, w
3represent normal condition.
Step C) traffic hazard data prediction: training sample set and above-mentioned N group historical data sample { X
1, X
2..., X
n, wherein, each X
i={ x
1, x
2..., x
11(I=1,2,3 ..., N) and N group casualty effect { Y corresponding thereto
1, Y
2..., Y
n, wherein each Y
i(I=1,2 ..., N) correspond to { w
1, w
2, w
3in a certain item.
Step D) utilize above-mentioned sample data to train multi-class support vector machine, draw accidents classification decision function
Parameter a
j, b
j.
Step e) to above-mentioned N group data sample classification, to sample X
i(I=1,2,3 ..., N), if
then X
iclassification be g (X
i)=w
j.
Step F) if sample X
iprediction classification g (X
i) be different from true classification Y
i, then optimize multi-class support vector machine parameter, go to step D), until reach best result class precision.
Step G) gather real time traffic data X every s second (0.01≤s≤0.1), i.e. the sex x of people
1, driving efficiency x
2; The spacing x of car
3, speed of a motor vehicle x
4, vehicle condition x
5; The type x of road
6, pavement behavior x
7, traffic x
8, and the weather x on the same day
9, pedestrian situation x
10, visibility x
11.According to criterion
x is assigned to w
jin class, i.e. classification g (the X)=w of X
j.
Step H) if g (X) equals w
1or w
2then differentiate that this section is current and have the risk that traffic hazard occurs, early warning is carried out to driver, as passed through variable message board in this front, section, and start opertaing device, by controlling the ring road of road or the Intersections of through street, reduce upstream vehicle flow, by variable speed-limit plate to Current vehicle speed limit, reduce the travel speed of upstream vehicle.
Step I) if g (X) equals w
3, then this section is current is safe condition, without the need to the prompting that gives the alarm, goes to step G).
Compared with prior art, technical scheme of the present invention has following beneficial effect:
Accident diagnoses accuracy rate is high.Existing street accidents risks detection method only utilizes traffic flow data to calculate traffic hazard probability, and the present invention acquires the several data information affecting traffic hazard simultaneously, thus can improve street accidents risks accuracy of detection.The traffic information data in the section of Real-time Collection is brought into the traffic hazard discriminant classification function after study, in real time traffic hazard is occurred to section and detect.Whether occur according to traffic hazard, determine current the need of startup early warning means, reduce street accidents risks, thus improve the accuracy rate of vehicle regulation and control, reduce traffic hazard, effectively ensured the traffic safety of through street.Testing process is simple.In the present invention, only need after obtaining classification function to gather the real time traffic data information in section, just can in real-time estimate setting-up time in future, whether this section there is traffic hazard, easy to use, practical, has good application prospect.
Accompanying drawing explanation
Fig. 1 is accident identification and classification FB(flow block) of the present invention.
Fig. 2 is the real-time testing process schematic diagram of road in the present invention.
Embodiment
Below in conjunction with drawings and Examples, technical scheme of the present invention is described in further detail.
The invention discloses a kind of based on the real-time traffic accident risk differentiation of support vector machine analysis road and control method, according to the historical traffic accident information gathered, the discriminant classification function utilizing multi-class support vector machine to set up, whether predicted detection section exists the risk that traffic hazard occurs, if occurred, take the corresponding opertaing device of Real-time Road, otherwise continue image data, judge next time; The advantage of the inventive method is that the real-time traffic characteristic parameter utilizing road traffic checkout equipment to obtain carries out real-time estimate to traffic hazard, and has good precision of prediction.Thus, this method is differentiating the risk of through street traffic hazard, and prediction traffic hazard generation aspect has actual engineering application and is worth.
The Real-time Traffic Information collected, weather data, pedestrian's data, vehicle data information are brought in the traffic hazard identification and classification function that the present invention sets up, whether calculate current generation traffic hazard.Have an accident if recorded, then showing that this section is current has the risk that traffic hazard occurs, and should give the alarm, and the control program activated in dynamic traffic control system reduces accident risk, then continues image data; Do not have an accident if recorded, then continue image data, judge next time.
Vehicle regulate and control method of the present invention, according to the real time traffic data, weather data, pedestrian's data, the vehicle data that gather, judges to detect whether section is current exists the risk that traffic hazard occurs, be adopt multi-class support vector machine analysis.
Practice process of the present invention is divided into be set up traffic hazard discriminant function relational expression and detects traffic hazard two processes.
Set up accidents classification discriminant function relational expression: collection or casualty data, weather data and traffic data in a period of time of acquisition testing section.In order to ensure that the accident forecast function set up can have good precision of prediction, the sample of collection is as far as possible large, and usual accident group data sample is greater than 200, and normal group data sample is greater than 400.According to above-mentioned steps A) to step F) draw identification and classification function by training sample set.
Claims (5)
1. a road real-time traffic accident risk control method, is characterized in that, concrete steps are:
Steps A) gather under normal traffic states and before having accident to occur t second (0.01≤t≤0.1) state under data sample N group { X
1, X
2..., X
n; Each X
i(I=1,2,3 ..., N) comprise people, car, road, environmental information: according to time of origin and the generation section of traffic hazard, there is the sex x of people from section in random acquisition traffic hazard
1, driving efficiency x
2; The spacing x of car
3, speed of a motor vehicle x
4, vehicle condition x
5, road type x
6, pavement behavior x
7, traffic x
8; And the weather x on the same day
9, pedestrian situation x
10, visibility x
11;
Step B) gather and above-mentioned data sample { X
1, X
2..., X
ncorresponding traffic hazard classification sample { Y
1, Y
2..., Y
n, each Y
i(I=1,2,3 ..., N) correspond to { w
1, w
2, w
3in a certain item, wherein w
1represent rear-end impact accident, w
2represent and hit fixture and stationary vehicle, w
3represent normal condition;
Step C) traffic hazard data prediction: training sample set and above-mentioned N group historical data sample { X
1, X
2..., X
n, wherein, each X
i={ x
1, x
2..., x
11(I=1,2,3 ..., N) and N group casualty effect { Y corresponding thereto
1, Y
2..., Y
n, wherein each Y
i(I=1,2 ..., N) correspond to { w
1, w
2, w
3in a certain item;
Step D) utilize above-mentioned sample data to train multi-class support vector machine, draw accidents classification decision function
Parameter a
j, b
j;
Step e) to above-mentioned N group data sample classification, to sample X
i(I=1,2,3 ..., N), if
then X
iclassification be g (X
i)=w
j;
Step F) if sample X
iprediction classification g (X
i) be different from true classification Y
i, then optimize multi-class support vector machine parameter, go to step D), until reach best result class precision;
Step G) gather real time traffic data X second every s, i.e. the sex x of people
1, driving efficiency x
2, car spacing x
3, speed of a motor vehicle x
4, vehicle condition x
5, road type x
6, pavement behavior x
7, traffic x
8, and the weather x on the same day
9, pedestrian situation x
10, visibility x
11, according to criterion
x is assigned to w
jin class, i.e. classification g (the X)=w of X
j;
Step H) if g (X) equals w
1or w
2, then differentiating that this section is current has the risk that traffic hazard occurs, and carries out early warning to driver,
Step I) if g (X) equals w
3, then this section is current is safe condition, without the need to the prompting that gives the alarm, goes to step G).
2. according to road real-time traffic accident risk control method according to claim 1, it is characterized in that, detection road installs Vehicle License Plate Recognition System, the corresponding information data of driver in data bank is transferred for each board, extract its sex and driving age and passing traffic hazard historical data.
3., according to road real-time traffic accident risk control method according to claim 1, it is characterized in that, described step G) in, transport information sampling time interval s meets 0.01≤s≤0.1.
4. according to road real-time traffic accident risk control method according to claim 1, it is characterized in that, described step H) in early warning be: as passed through variable message board in this front, section, and start opertaing device, by controlling the ring road of road or the Intersections of through street, reduce upstream vehicle flow, by variable speed-limit plate to Current vehicle speed limit, reduce the travel speed of upstream vehicle.
5. according to road real-time traffic accident risk control method according to claim 4, it is characterized in that, by variable speed-limit plate to Current vehicle speed limit, the car speed amplitude of each adjustment change is within 5km/h.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310596435.2A CN103646534B (en) | 2013-11-22 | 2013-11-22 | A kind of road real-time traffic accident risk control method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310596435.2A CN103646534B (en) | 2013-11-22 | 2013-11-22 | A kind of road real-time traffic accident risk control method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN103646534A CN103646534A (en) | 2014-03-19 |
CN103646534B true CN103646534B (en) | 2015-12-02 |
Family
ID=50251741
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201310596435.2A Expired - Fee Related CN103646534B (en) | 2013-11-22 | 2013-11-22 | A kind of road real-time traffic accident risk control method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103646534B (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106095963A (en) * | 2016-06-17 | 2016-11-09 | 上海经达信息科技股份有限公司 | Vehicle drive behavior analysis big data public service platform under the Internet+epoch |
CN111859291A (en) * | 2020-06-23 | 2020-10-30 | 北京百度网讯科技有限公司 | Traffic accident recognition method, device, equipment and computer storage medium |
Families Citing this family (29)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US9519670B2 (en) * | 2014-08-29 | 2016-12-13 | Ford Global Technologies, Llc | Method and apparatus for road risk indices generation |
CN106491144B (en) * | 2016-09-22 | 2019-07-05 | 昆明理工大学 | A kind of test and evaluation method of the latent risk perceptions ability of driver |
CN108154681B (en) * | 2016-12-06 | 2020-11-20 | 杭州海康威视数字技术股份有限公司 | Method, device and system for predicting risk of traffic accident |
CN106651162A (en) * | 2016-12-09 | 2017-05-10 | 思建科技有限公司 | Big data-based driving risk assessment method |
CN106485922B (en) * | 2016-12-20 | 2019-03-12 | 东南大学 | Secondary traffic accident method for early warning based on high-precision traffic flow data |
CN109146217A (en) | 2017-06-19 | 2019-01-04 | 北京嘀嘀无限科技发展有限公司 | Safety travel appraisal procedure, device, server, computer readable storage medium |
CN107978149A (en) * | 2017-11-17 | 2018-05-01 | 嘉兴四维智城信息科技有限公司 | Typhoon weather urban traffic accident probabilistic forecasting processing unit and its method |
CN107942411B (en) * | 2017-11-30 | 2020-04-17 | 南京理工大学 | Atmospheric visibility prediction method |
US10235882B1 (en) * | 2018-03-19 | 2019-03-19 | Derq Inc. | Early warning and collision avoidance |
CN108710967B (en) * | 2018-04-19 | 2021-07-27 | 东南大学 | Expressway traffic accident severity prediction method based on data fusion and support vector machine |
CN109285344B (en) * | 2018-06-25 | 2021-05-28 | 江苏智通交通科技有限公司 | Identification method and intelligent decision-making system for key monitoring objects of high-risk traffic personnel |
CN108960501B (en) * | 2018-06-28 | 2021-11-19 | 上海透云物联网科技有限公司 | Commodity anti-channel conflict method |
CN110717035A (en) * | 2018-07-11 | 2020-01-21 | 北京嘀嘀无限科技发展有限公司 | Accident rapid processing method, system and computer readable medium |
CN109117987B (en) * | 2018-07-18 | 2020-09-29 | 厦门大学 | Personalized traffic accident risk prediction recommendation method based on deep learning |
CN110858334A (en) * | 2018-08-13 | 2020-03-03 | 北京中科蓝图科技有限公司 | Road safety assessment method and device and road safety early warning system |
CN109410567B (en) * | 2018-09-03 | 2021-10-12 | 江苏大学 | Intelligent analysis system and method for accident-prone road based on Internet of vehicles |
CN109300310B (en) * | 2018-11-26 | 2021-09-17 | 平安科技(深圳)有限公司 | Traffic flow prediction method and device |
CN109710984A (en) * | 2018-12-04 | 2019-05-03 | 斑马网络技术有限公司 | Identification of accidental events and rescue mode and device |
CN110033386B (en) * | 2019-03-07 | 2020-10-02 | 阿里巴巴集团控股有限公司 | Vehicle accident identification method and device and electronic equipment |
JP2022546320A (en) | 2019-08-29 | 2022-11-04 | ディーイーアールキュー インコーポレイテッド | Advanced in-vehicle equipment |
US11625624B2 (en) | 2019-09-24 | 2023-04-11 | Ford Global Technologies, Llc | Vehicle-to-everything (V2X)-based real-time vehicular incident risk prediction |
CN111680897A (en) * | 2020-05-27 | 2020-09-18 | 公安部道路交通安全研究中心 | Service management and control method, device and system and computer readable storage medium |
CN112365162B (en) * | 2020-11-12 | 2024-03-08 | 北京交通大学 | Railway operation risk control method based on accident cause network |
CN112562337B (en) * | 2020-12-10 | 2022-05-13 | 之江实验室 | Expressway real-time traffic accident risk assessment method based on deep learning |
CN112784723A (en) * | 2021-01-14 | 2021-05-11 | 金陵科技学院 | Road traffic safety protection model based on IFast-RCNN algorithm |
CN113256014B (en) * | 2021-06-02 | 2021-10-22 | 沸蓝建设咨询有限公司 | Intelligent detection system for 5G communication engineering |
CN113744526B (en) * | 2021-08-25 | 2022-12-23 | 贵州黔通智联科技股份有限公司 | Highway risk prediction method based on LSTM and BF |
CN115359655A (en) * | 2022-08-03 | 2022-11-18 | 中远海运科技股份有限公司 | Highway monitoring system based on hierarchical clustering algorithm |
CN116403403B (en) * | 2023-04-12 | 2024-02-02 | 西藏金采科技股份有限公司 | Traffic early warning method, system, equipment and medium based on big data analysis |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102298706A (en) * | 2011-08-12 | 2011-12-28 | 河海大学 | Inland waterway ship large-scale prediction method in restricted conditions |
Family Cites Families (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP4940206B2 (en) * | 2008-09-09 | 2012-05-30 | 株式会社東芝 | Road traffic information providing system and method |
JP5398522B2 (en) * | 2009-12-29 | 2014-01-29 | 株式会社東芝 | Road traffic control support information creation system |
EP2648133A1 (en) * | 2012-04-04 | 2013-10-09 | Biomerieux | Identification of microorganisms by structured classification and spectrometry |
-
2013
- 2013-11-22 CN CN201310596435.2A patent/CN103646534B/en not_active Expired - Fee Related
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102298706A (en) * | 2011-08-12 | 2011-12-28 | 河海大学 | Inland waterway ship large-scale prediction method in restricted conditions |
Non-Patent Citations (2)
Title |
---|
城市路网交通预测模型研究及应用;张扬;《中国博士学位论文全文数据库工程科技Ⅱ辑》;20130315;C034-15 * |
道路交通事故预测及控制研究;雷兢;《中国优秀硕士学位论文全文数据库工程科技Ⅱ辑》;20050615;C034-240 * |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106095963A (en) * | 2016-06-17 | 2016-11-09 | 上海经达信息科技股份有限公司 | Vehicle drive behavior analysis big data public service platform under the Internet+epoch |
CN111859291A (en) * | 2020-06-23 | 2020-10-30 | 北京百度网讯科技有限公司 | Traffic accident recognition method, device, equipment and computer storage medium |
CN111859291B (en) * | 2020-06-23 | 2022-02-25 | 北京百度网讯科技有限公司 | Traffic accident recognition method, device, equipment and computer storage medium |
Also Published As
Publication number | Publication date |
---|---|
CN103646534A (en) | 2014-03-19 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103646534B (en) | A kind of road real-time traffic accident risk control method | |
CN102360525B (en) | Discriminant analysis-based high road real-time traffic accident risk forecasting method | |
CN102360526B (en) | Real-time monitoring method for road section state of high road | |
CN102568226B (en) | High speed variable speed limit control method based on adverse weather conditions | |
CN104392610B (en) | Expressway based on distributed video traffic events coverage dynamic monitoring and controlling method | |
Yang et al. | Estimation of traffic conflict risk for merging vehicles on highway merge section | |
CN103914688A (en) | Urban road obstacle recognition system | |
CN105946860B (en) | A kind of bend speed prediction method for considering driving style | |
CN105118316A (en) | Curved road safe speed calculating method and caution system based on vehicle infrastructure cooperation | |
CN103971523A (en) | Mountainous road traffic safety dynamic early-warning system | |
CN104992145A (en) | Moment sampling lane tracking detection method | |
CN103198713A (en) | Traffic accident reduction vehicle regulation and control method based on traffic data and weather data | |
CN103473928A (en) | Urban traffic jam distinguishing method based on RFID technology | |
CN104008644B (en) | A kind of traffic noise on urban roads measuring method based on Gradient Descent | |
CN103606269A (en) | Control method for improving traffic efficiency of freeway construction area | |
CN102568206A (en) | Video monitoring-based method for detecting cars parking against regulations | |
CN108597219A (en) | A kind of section pedestrian's street crossing control method based on machine vision | |
CN202422420U (en) | Illegal parking detection system based on video monitoring | |
CN103198707B (en) | A kind of vehicle regulate and control method based on traffic flow character dangerous under fine day situation | |
CN107784832A (en) | A kind of method and apparatus for being used to identify the accident black-spot in traffic route | |
Goh et al. | Experimental microsimulation modeling of road safety impacts of bus priority | |
CN102360524B (en) | Automatic detection and confirmation method of dangerous traffic flow characteristics of highway | |
CN100481153C (en) | Method for automatically inspecting highway traffic event based on offset minimum binary theory | |
CN109050521A (en) | A kind of expressway bend rollover sideslip early warning system and method | |
CN106530714A (en) | Secondary traffic accident time prediction method based on traffic flow data |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | 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 |
Granted publication date: 20151202 Termination date: 20161122 |
|
CF01 | Termination of patent right due to non-payment of annual fee |