Parameter Calibration for VISSIM Using a Hybrid Heuristic Algorithm: A Case Study of Congested Traffic Networks in China

This paper presents an application of a hybrid heuristic algorithm as an optimization method to find a suitable combination of micro-simulation parameters. The calibration is based on field data collected on weekday peak time of an urban downtown road network in Hefei, China. The tests indicated that the hybrid algorithm can well combine the advantage of Genetic Algorithm (GA) and Simulated Annealing (SA) algorithm. It can enhance the efficiency of convergence, and effectively reduce the likelihood of premature and converge to local optima. Meanwhile, eight important parameters and their proposed value were found which will affect the simulation accuracy especially when modeling the local peak time traffic conditions.