Sorting Model of Optimization Order in Traffic Signal Planning

Abstract

In optimizing the signal timings of a network, determining the sequence in which the intersections will be optimized is an important issue. Determining the critical intersection in the network is another important problem. Optimization order and the critical intersection are determined by users on the basis of experience or by a simple greedy method such as the saturation degree sorting method. However, the signal timing plan that is generated may not be the true optimal plan. This issue is addressed by a new method, the Sorting Model of Optimization Order (SMOO), for finding the optimal order for optimization and the critical intersection. The intersection in first position on the optimization order list is defined as the critical intersection. Two cases are used to test the model. In experiments, SMOO computes an order that is closer to optimal than those of other greedy sorting methods. According to measures of effectiveness, the signal plans produced by SMOO outperform those produced by other methods.