【作者】 曾松林；

【导师】 余立建；

【作者基本信息】 西南交通大学 ， 交通信息工程及控制， 2013， 硕士

【摘要】 近年来,随着国民经济的高速发展及城市化进程的推进,交通日趋拥挤,交通事故频繁,传统的控制方法己经不能有效地解决日益严重的交通问题。为此,需要一些新的技术和手段从根本上来解决交通问题,智能交通系统的出现为解决这一问题提供了可能。以城市单交叉口的信号控制为研究对象,利用模糊逻辑、神经网络以及粒子群优化算法等智能控制方法,对城市交通信号控制展开研究,以缓解城市交通拥堵,提高交通运输效益。对城市单交叉口多相位的经典模糊控制进行了分析研究,在其基础上综合考虑城市交通流的特点设计了一种相序可优化的模糊控制策略。该策略以交叉口车辆的平均延误作为信号控制的性能评价指标,综合考虑车辆的排队长度、红灯等待时间以及平均达到率等因素,以此来决定相位的切换顺序和绿灯分配时间。仿真结果验证了可变相序模糊控制的有效性。针对可变相序模糊控制器存在的缺陷,设计了一种基于粒子群算法的模糊控制方法。由于可变相序模糊控制器的隶属度函数和模糊规则是基于专家的经验而设计的,具有较大的主观性和随意性。本文利用粒子群算法对模糊控制器的隶属度函数和模糊规则进行优化,使之能够适应不同的交通流情况,从而实现对交叉口信号的自适应控制。仿真结果表明利用粒子群算法优化后的可变相序模糊控制能更有效地降低交叉口车辆的平均延误,进一步提高模糊控制器的性能。从基于交通流数据的角度出发,利用模糊神经网络对城市交叉口的信号控制进行建模。模糊神经网络由模糊逻辑和神经网络有机结合而成,吸取了二者的长处,弥补了各自的不足,有效地提高了整个系统对知识的学习和表达能力。仿真结果表明将其应用于交叉口的信号控制具有更大的优势。更多还原

【Abstract】 In recent years, with the rapid development of national economy and the acceleration of urbanization, traffic is becoming more and more crowded and the traffic accidents happen frequently. Traditional control methods can not effectively solve the increasingly serious traffic problems. Therefore, it needs some the latest technologies and means to fundamentally solve the traffic problems, and the emergence of intelligent transport system provides the possibility to solve this problem.The thesis takes the signal control of urban single intersection as the research object, some intelligent control methods such as fuzzy logic, neural network and particle swarm optimization algorithm are studied to alleviate the city traffic congestion and improve the efficiency of transportation.The classical fuzzy control on urban multi-phase single intersection is studied. Based on the comprehensive consideration of the characteristics of urban traffic flow, a fuzzy control strategy is presented in which the phase sequence can be optimized. The strategy takes the average delay of vehicles at the intersection as the index for evaluating the performance of signal control. In this strategy, various factors such as the queue length, the waiting time as well as the average arrival rate of vehicles are considered to determine the switching sequence of phase and distribution of green light time. The simulation results verify the effectiveness of the strategy.Aiming at the existing defects of the phase sequence-changeable fuzzy controller, a fuzzy control method is designed based on particle swarm optimization. Because the membership functions and fuzzy rules of the phase sequence-changeable fuzzy controller are designed based on the experience of experts, it has considerable subjectivity and arbitrariness. In this paper, the particle swarm optimization algorithm is used to optimize the membership functions and rules of the fuzzy controller so as to adapt to different traffic flow and realize the adaptive control of the traffic signal. The simulation results show that the new phase sequence-changeable fuzzy control method based on particle swarm optimization algorithm can efficiently reduce the average delay of vehicles at the intersection and further improve the control performance the fuzzy controller.From the perspective of traffic flow data, the fuzzy neural network is used for building the model of signal control for urban single intersection. Fuzzy neural network is organically constituted by fuzzy logic and neural network, which draw on their strengths and makes up for their shortcomings. It effectively improves the learning and expression ability of knowledge of the entire system. The simulation results show that its application in traffic signal control has more advantages.更多还原