Multi-objective Optimization Method of Fixed-Time Signal Control of Isolated Intersections

In this paper, we presented a multi-objective optimization model of fixed-time signal control of unsaturated intersections. The performance indices taken into consideration include total volume, average delay, average stop frequency, average queue length of vehicular traffic at signalized isolated intersections. A multi-objective genetic algorithm was given to solve the model based on Non-dominated Sorting Genetic Algorithm II (NSGAII) which can solve multi-objective optimization problems directly. Finally, an example was given to illustrate not only advantages of multi-objective optimization over single-objective or weighted combination optimization methods but also validities of different performance indices and sensitivities of the performance indices to signal control parameters. Compared with single-objective optimization method, multi-objective optimization method can sharply decrease average vehicle delay and queue length at a single intersection, and compared with weighted combination method, the multi-objective optimization method can acquire better comprehensive traffic benefits.