Comprehensive Evaluation And Bottleneck Problem Analysis For The Traffic Safety Of Freeway

With the increase of the freeway mileage, the freeway traffic safety situation has become increasingly severe in China and the traffic fatal-accident rate has been the highest in the world. Freeway accident not only endangers the safety of the people and property severely, but also restricts the using performance of freeway. So, it is urgently important subjects to scientifically evaluate freeway traffic safety situation, analyze the traffic safety bottleneck problem and find the improvement strategy now.The paper bases on the field survey of the traffic safety situation, traffic facilities and traffic situation, and combines the traffic safety's related achievement, uses the analytic hierarchic process, matter-element evaluation method, support vector machine (SVM) and factor analysis to evaluate the traffic safety situation and analyze the bottleneck problem. The chief contents include:Firstly, based on the analysis of present situation of the traffic safety, it discuss the influence factor of the freeway traffic safety and its interaction of the people, car, road, traffic environment and traffic management from the traffic engineering.Secondly, by synthesize the freeway traffic safety factors, it establish comprehensive evaluation index system of the freeway traffic safety based on the selection principles and grade scale of the evaluation indexes. Use the analytic hierarchic process to confirm the weight of the indexes and the matter-element evaluation method to select the train samples, establish a comprehensive evaluation model of the freeway traffic safety based on the support vector machine. Select the partial binary tree SVM multi-classification algorithm and use the matrix similarity degree method to confirm the SVM's parameters, Using MATLAB solving the comprehensive evaluation model, train and test the SVM's samples. The test result verifies the SVM comprehensive evaluation model's effectiveness.Thirdly, seeking the cumulative frequency curve method suited for the freeway black spot identification combined the statistical characteristics of the accident frequency. After finding the black spot, analyze the factors form the black spot. Then, use the factor analysis method to confirm the main causes and secondary causes of the black spot's and find out the reasons of the bottleneck problem. According to the bottlenecks problem governance principles, analyze the technology governance strategy of the freeway's traffic safety bottleneck problem.Finally, the paper takes the traffic safety of the Hukun Freeway (Changzhang section) for empirical research. Analyze the total traffic safety situation of the Changzhang section and mine its accident distribution character. Use the SVM comprehensive evaluation model to evaluate the Changzhang section traffic safety, the cumulative frequency curve method to find its black spots, factor analysis method to analyze the causes of the black spot and locate the bottleneck problem. Propose Changzhang Freeway safety improvement strategy according to result of the comprehensive evaluation and bottleneck problem analysis. The empirical research shows the feasibility and practicality of the SVM comprehensive evaluation model and bottleneck analysis method, so it can provide theoretical basis to establish the theoretical framework coordinates the people, vehicles, roads, traffic environment and traffic management and to improve the safety level of freeway.