Analysis And Research On Intelligent Control Algorithm Of Urban Area Traffic Signal

In recent years, with the social and economic development and urbanization process is accelerating, and the increasing of urban population and vehicle, urban traffic congestion has become more serious, and gradually become the one important factor restricting the development of urban economy. The reasonable control of urban area traffic signal can effectively alleviate traffic congestion, and traditional control method is difficult to apply in the complicated traffic control problem. Therefore, it is very important to study the intelligent control method of urban area traffic signal.Using neural network as the main method, and combining with the particle swarm optimization algorithm and fuzzy control and other intelligent algorithms, to study the signal control of urban area.The city intersection neural network self-learning control is analyzed, based on the single intersection neural network self learning control, a coordinated control method of regional signal based on BP neural network is presented. The method retains the self learning structure of neural network, and considers the number of vehicles arriving at the adjacent intersections. The simulation results verify the effectiveness of the coordinated control method of regional signal based on BP neural network.In view of the BP neural network is easy to fall into local optimal solution and slow training speed defects, the particle swarm neural network and simulated annealing particle swarm neural network are applied to the coordinated control method of regional signal based on neural network. Simulation results show that the control performance of the coordinated control method of regional signal based on particle swarm neural network or simulated annealing particle swarm neural network is improved, which, the control effect of the coordinated control method of regional signal based on simulated annealing particle swarm neural network is better.Aiming at the cyclical adjustment mode’s defects of the coordinated control method of regional signal based on neural network and other problems, combined with the fuzzy control and the coordinated control method of regional signal based on neural network, a fuzzy and neural network hybrid control method is designed. The simulation results show that the fuzzy and neural network hybrid control method can achieve better control effect.