Signal Optimization Control for Regional Intersection Based on Switching Fuzzy System

By taking a single intersection at the key node of non-equilibrium road network as the research object, a multistate combined intersection signal optimization control method based on fuzzy switching control was proposed from two aspects of timing modeling optimization and multi-state switching control. The self-learning algorithm was utilized to divide the traffic state of each phase into four states: low traffic flow, medium traffic flow, high traffic flow and saturated traffic flow. Combined switching control strategy was introduced to select the appropriate control model in different traffic conditions. As indicated by the simulation results, the combined switching optimization control method boasts of improving the deficiency of the two-stage fuzzy controller in increasing the average stopping times in low traffic flow, and effectively reducing the average delay of vehicles in medium and high traffic flow. ICGA can improve the search efficiency, and is superior to SGA and CGA in performance. It has achieved favorable results in signal optimization control of non-equilibrium intersections in key sections.