A computational experiment method in ACP framework for complex urban traffic networks

Urban traffic congestion has already become an urgent problem. Artificial societies, Computational experiments, and Parallel execution (ACP) method is applied to urban traffic problems. In ACP framework, optimization for urban road networks achieves remarkable effect. Optimization for urban road networks is a problem of nonlinear and non-convex programming with typical large-scale continual and integer variables. Due to the complicated urban traffic system, this paper focuses on the ACP-based Computational experiments modeling. It hopes to find an optimization model that is further accord with the practical situation. To this end, we use a mixed integer nonlinear programming problem (MINLP) and an genetic algorithm (GA) for urban road networks optimization. The systemic simulation experiments show that the approach is more effective in improving traffic status and increasing traffic safety.