Study on Fuzzy Neural Network Model of Vehicle-Bicycle Conflicts at Grade Intersection Based on Failure Tree Analysis

摘要

The bicycle safety problems at urban grade intersection should be regarded in China. Firstly, according to the amount of conflicts and the degree of danger, the vehicle-bicycle conflicts at a four-leg intersection are classified into 3 groups using the fuzzy clustering theory, and the interference between the through bicycles and the right-turn vehicles in the adjacent lane is used as the case study. The failure tree model of bicycle traffic safety is built and the main reasons for vehicle-bicycle conflict are explained from two aspects: the minimal cut set and the structural importance. Secondly, based on the failure tree analysis results, the fuzzy neural network model of vehicle-bicycle conflicts at grade intersections is built, the analysis results show that the reasons causing bicycle traffic conflict at grade intersections can be well analyzed and evaluated using the fuzzy neural network model.