Parameter Correction of VISSIM Multi-intersection Simulation Model Based on Combined Genetic Algorithm

Abstract

In the field of traffic simulation research, the requirements of microscopic traffic simulation model accuracy and intersection popularity are getting higher and higher. Aiming at the shortcomings of traditional genetic algorithm, such as slow speed, huge time-consuming and easy to fall into local optimum, a parameter correction algorithm for VISSIM simulation model based on combined genetic algorithm is proposed in this paper. Neural network is used to predict and analyze, and genetic algorithm with partitioning operator is used to correct the parameters, and the error of simulation delay and measured delay is designed as the objective function. Finally, taking three typical continuous intersections of a main line as an example, a microscopic traffic simulation model is established and the parameters are corrected. It is showed by the results that the corrected error is as low as 9.63%, which proves that the proposed method is feasible and robust.