【摘要】 延安市宝塔区杨家岭路段交叉口是延安市城区内的重要交叉口，近年来由于城市交通流量的日益增加，该交叉口的交通拥堵问题日益严重，传统的信号配时方法已不能满足实际需求，需要新的信号配时优化方法和模型来缓解该交叉口的交通通行压力.为解决此问题本研究首先分析杨家岭路段十字交叉口存在的问题，进而采用多目标粒子群算法和GRU循环神经网络建立模型，最后利用VISSIM软件分别对两相位交叉口和多相位交叉口进行仿真实验验证，结果反映了基于MOPSO-GRU神经网络的信号配时优化模型可提高交通流量的效率和整体交通系统的性能.更多还原

【Abstract】 The intersection of Yangjialing section in Baota District of Yan’an City is an important intersection in the urban area of Yan’an City. In recent years, due to the increasing urban traffic flow, the traffic congestion problem at the intersection has become increasingly serious. The traditional signal timing method can no longer effectively meet the actual needs. New signal timing optimization methods and models are needed to alleviate the traffic pressure at the intersection. In order to solve this problem, this study first analyzes the problems existing in the cross intersection of Yangjialing section, and then uses the multi-objective particle swarm optimization algorithm and GRU recurrent neural network to establish the model. Finally, VISSIM software is used to simulate the two-phase intersection and multi-phase intersection respectively. The simulation results show that the signal timing optimization model based on MOPSO-GRU neural network can improve the efficiency of traffic flow and the performance of the overall traffic system.更多还原