Agent simulation model of right-turning motor vehicles crossing through bicycles at signalized intersection based on BP neural network

Firstly, according to behavior analysis of right-turning motor vehicles crossing through bicycles at signalized intersection, a BP neural network decision model of vehicles-Agent is set up. Then, an Agent simulation model is built by the signalized intersection structure analysis. Finally, the simulation results of the typical intersection in Beijing show that the motor judgment ability on crossing time is accurate and intersection traffic capacity is improved, which indicates that the Agent simulation model based on BP neural network vehicles-Agent decision model is practical and effective.