|
With the freeway mileage and traffic growth, Chinese freeway accident indicators have continued to grow, this causes huge losses for the state and people. Therefore, to reduce freeway traffic accident victims or lower level is an urgent problem. Microscopic evaluation of freeway traffic safety focuses on the relationship between traffic accidents and environmental factors. Analyzing specific reasons causing traffic accidents from different points of view, to improve freeway traffic safety has important practical significance.On the basis of existing research to evaluation of traffic safety at home and abroad, the paper comprehensively analyzes the effects of traffic safety from the elements of freeway transportation system as the starting point, respectively, from people, vehicles, roads, environment and management. And following the principles of evaluation indicator system construction, a multi-layered microscopic evaluation index system of freeway traffic safety has been established. Microscopic evaluation method of freeway traffic safety based on common methods of advantages, disadvantages and application characteristic is determined the BP network evaluation. In this paper, genetic algorithm and BP network have combined perfected with MATLAB software to create a high precision, good performance of GA-BP network model. The model has been trained with samples whose safety levels are clear. Then, the instance has been evaluated using the trained model, and the evaluation result is the same as the actual safety level of the freeway. Finally, analyze the evaluation results and the effective measures are proposed to the poor safety level of freeway.
|