Traffic Control And Guidance System Research Based On Evolutionary Multi-objective Optimization And Ant Algorithm

With the increasing of traffic flow and complication,city road network congestion problem becomes more and more serious. The existing intelligent traffic control is difficult to improve the overall efficiency of city traffic system. It is an effective way to solve urban traffic problems with guidance system, which can balance the traffic flow initiatively, and control system, which can relieve traffic flow passively. How to build such optimization model of intelligent transportation system and optimization algorithm becomes the current research focus and key technology. For existing traffic control system problems, this dissertation adopts a the evolutionary multi-objective optimization control system, combining with a guidance system based on ant algorithm. This dissertation constructs a multi-objective optimization of single intersection control model, multi-intersection coordination mechanism and vehicle guidance system, which can effectively balance the traffic load and improve traffic efficiency metropolitan road network. The main work is as follows:1) For existing traffic control system can not effectively take a variety of indicators into account, and can’t efficiently regulate control signal timing plans based on real-time traffic conditions, this dissertation constructs a single intersection multi-objective optimal control model, which is optimized by the improved evolutionary multi-objective optimization algorithm. This model bases on the number of vehicles which through the intersection per unit of time and the average time of vehicles’ delay each cycle. This model could adjust their timing plan based on real-time road traffic information, and provide transportation policy makers timing plan of various kinds of bias. In order to meet the requirement of traffic control system based on multi-objective optimization, this dissertation proposes a parallel non-dominated sorting multi-objective optimization algorithm. Simulation shows that the algorithm has higher timeliness, strong pareto frontier for exploration capability and ability to maintain population diversity.2) Existing regional multi-intersection coordination has a high coupling degree with single intersection signal control system, which makes the intersection coordination become complexity, a bad real-time performance, and low forecasting ability in theintersection congestion issue. This dissertation constructs a multi-intersection coordination mechanism, which adjusts the ratio of the real traffic flow and the saturated traffic flow. According to the characteristics of the coordination mechanism, this dissertation uses fuzzy control technology to simulate. Simulation shows that the coordination mechanism reduces the response time of the traffic congestion, coordinate signal control all the intersections quickly,which improves regional transportation efficiency.3) For existing guidance system less considers dynamic price on the road, starting point and end destination of travelers, this dissertation constructs a vehicle guidance model based on a variety of indicators, which is optimized by the improved ant algorithm. This optimization system’s goal is consisted of three aspects: the static cost between origin and destination, the cost through the intersection, the dynamic cost of running on the road. This model keeps the users’ optimal path and tries to achieve a balanced distribution of the vehicles in road. In order to meet the requirements of path planning guidance system, this dissertation proposes a preferent ant algorithm. This algorithm has a high global search ability and efficiency by setting preferences and jumping out of local optimal mechanism. Simulation shows that the algorithm has high performance for guidance system.