Study On The Method Of Urban Road Traffic Safety Risk Monitoring

With the acceleration of urbanization,the urban road traffic safety management has posed a severe challenge.The possibility of traffic accidents can be expressed with traffic safety risk.In order to improve the level of urban road traffic safety,reduce the urban road traffic accident probability,it is necessary to carry out real-time monitoring on urban road traffic safety risks.Firstly,the method of twice divisions in the urban road network had been proposed to realize the purpose of real-time monitoring.In the first division,each minimum grid of the urban road network had been defined as a sub-area of the road network according to the physical segmentation of the urban road network.However,the intersection had different attributes such as signal light control and the number of lanes.If the components of the road network sub-area had been directly integrated,the high-risk area would be diluted by the low-risk area,thus reduced the risk of the road network sub-area.Therefore,in the second division,the concept of "road network sub-domain" had been proposed according to the properties of intersections.Secondly,the division principle of key risk sources had been clarified,the designing principle of monitoring indicators had been determined,the monitoring indexes of key risk sources in the sub-domain of the road network had been defined,and then the three-level index monitoring architecture of urban road network—urban road network sub-area—urban road network subdomain had been established;Through the comparison of subjective and objective weighting methods,and with the application of information entropy theory,the information entropy weighting method had been applied to index weighting.Thirdly,the two division of urban road network must correspond to the two fusion.In the first fusion,the risks at intersections,sections and blocks of the road network sub-domain had been fused to obtain the risk level of the road network sub-domain.With the application of information fusion technology as the framework,the evidence correlation coefficient had been used to measure the conflict between evidences.Fuzzy set theory had been introduced to classify the risks of monitoring indicators.The risk early-warning system with indicators of 10 risk levels,risk classifications of 5 grades and warning classifications of 3 grades had been established.In the second fusion,the risks of the sub-domain of the road network had been fused,and the maximum risk of the sub-domain had been taken as the risk level of the sub-area of the road network.The risk monitoring model of urban road traffic safety had been constructed by these two fusions.Finally,the urban road network in Longgang district of Shenzhen city had been taken as an example,three types of sub-domains of the road network had been selected in the field,and two groups of experiments had been carried out.In experiment 1,the locations had been taken as thevariable to monitor the real-time risk changes of three types of sub-domain of the road network within 24 hours of the same day.The results showed that the risk levels of the road network's sub-domain with different attributes could be identify by the method of evidence correlation algorithm.In experiment 2,a seed domain had been randomly selected with time as a variable to monitor the daily variation of risk in the same period of two consecutive months.The results showed that the risk status of the same sub-domain of the road network could be distinguished by the method of evidence correlation algorithm in the same period of time with different dates.