A Golden Ratio-Based Genetic Algorithm and Its Application in Traffic Signal Timing Optimization for Urban Signalized Intersections

Abstract

Aiming at the problems of the local optimum of the genetic algorithm in solving traffic signal timing models, this paper presents an optimized traffic signal controller with a golden ratio-based genetic algorithm (TSCGRGA) for urban signalized intersections. This controller analyzes the correlation between the traditional performance indexes of traffic signal benefit, select time delays, stop rate, and throughput to establish relative evaluation indices of traffic signals; and a multiobject timing model of traffic signals is developed by weighted coefficient method. Then, an adaptive golden ratio-based genetic algorithm is presented to solve the optimized models, which uses real number encoding and introduces a golden ratio calculator to enhance the local optimal capability of genetic algorithm. Experiments are conducted on a typical urban isolated intersection and the performance of the developed model and algorithm is validated by comparison to those of fixed-time, actuated, and the real-coded genetic algorithm-based controllers for different traffic conditions. Extensive numerical calculation results have demonstrated the potential of the developed algorithm in the quality of solutions and time efficiency and indicated that the signal strategies derived from TSCGRGA have better performance.