Two-Stage Fuzzy Control for Intersections Based on Genetic Algorithm

The control rules on general fuzzy controller depend on expert's experiences, which have fixed quantitative factors and can't dynamically changes with the traffic flow change. In order to solve this problem, we have designed a two-stage fuzzy control system based on genetic algorithm for urban traffic intersections. The first-stage is phase selection module, the second-stage is green delay module. On the basis of the control decision-making made by the fuzzy controller, we use genetic algorithms (GA) to adjust the values on the fuzzy control query table coming from the output of fuzzy controller and optimize its values by using the average pass rate of vehicles as objective function. Through a simulation for a four-phase intersection, the result shows that this control algorithm has many advantages including flexible combination of phase, big average pass rate of vehicles, and it can effectively solve the problem of imbalance between traffic flow. So it is more suitable for real-time variable traffic conditions at urban intersections.