基于模糊逻辑的城市交通信号优化控制
发布时间：2018-07-15 16:50
【摘要】：随着机动车辆的迅速增加,城市交通拥堵问题日益加剧,制约着经济社会的发展,给居民的出行带来了很大的不变,解决城市交通拥堵变得尤为重要。提升交通信号的控制效率已成为解决交通拥堵的一种有效方法。本文针对目前交叉口信号控制方法和控制子区划分方法的不足,结合智能控制理论,在交叉口信号控制方案的选择、交叉口重要性评估、控制子区的动态划分等方面进行了探讨分析,具体的研究内容如下:(1)系统地介绍了交通信号控制理论和模糊控制理论,在交叉口信号控制中分别采用了单级模糊控制器和两级模糊控制器,研究了不同交通流下两种控制器的控制效果,通过分析控制器输入变量的选择,确定了不同交通流下控制器结构的配置。(2)针对在单交叉口信号控制中控制模型单一导致不能很好适应交通流变化的问题,提出了一种将单级模糊控制和两级模糊控制两种策略相结合的组合控制模型,并利用SOM神经网络实现了两种控制策略的切换。针对两级模糊控制器控制规则和隶属度函数人工设定的不足,利用混沌遗传算法优化两级模糊控制器参数。同时为了保证控制的实时性,结合滑动时间窗,根据实时采集数据实现交通状态的快速识别和两级模糊控制器参数在线优化。通过实例仿真,对提出的组合优化控制模型进行了验证。(3)以交叉口连接度、高峰车流量和车道占有率作为评价指标,应用熵权TOPSIS法对交叉口的重要性进行分析,给出了交叉口重要性评价的具体步骤,实现了城市区域路网的关键节点选择。(4)针对多数交通控制子区划分方法只考虑两交叉口关联度而忽略了关键交叉口的重要性的问题,提出了基于关键交叉口交通控制子区的动态划分方法,以关键交叉口为起点遍历四周交叉口,利用模糊推理求得两相邻交叉口之间的关联度,在此基础上,通过计算协调交叉口与关键交叉口的周期差确定控制子区的划分,并结合实例对提出的子区划分方法进行了分析研究。
[Abstract]:With the rapid increase of motor vehicles, the problem of urban traffic congestion is becoming more and more serious, which restricts the development of economy and society, and brings great invariance to the travel of residents, so it becomes more and more important to solve the urban traffic congestion. Improving the control efficiency of traffic signals has become an effective method to solve traffic congestion. In view of the deficiency of the current signal control method and the control sub-area division method of intersection, combining with the intelligent control theory, the selection of signal control scheme and the importance evaluation of intersection are discussed in this paper. The dynamic division of the control sub-area is discussed and analyzed. The specific research contents are as follows: (1) the traffic signal control theory and fuzzy control theory are introduced systematically. The single stage fuzzy controller and two stage fuzzy controller are used in the intersection signal control. The control effect of the two controllers under different traffic flow is studied, and the selection of the input variables of the controller is analyzed. The configuration of controller structure under different traffic flow is determined. (2) aiming at the problem of single control model in single intersection signal control, it can not adapt to traffic flow change well. This paper presents a combined control model which combines single-stage fuzzy control and two-level fuzzy control, and realizes the switching of the two control strategies using SOM neural network. Aiming at the deficiency of manual setting of control rules and membership function of two-stage fuzzy controller, the parameters of two-stage fuzzy controller are optimized by chaos genetic algorithm. At the same time, in order to ensure the real-time control, combined with sliding time window, according to the real-time data acquisition, the fast identification of traffic state and the on-line optimization of two-stage fuzzy controller parameters are realized. Through the example simulation, the combined optimal control model is verified. (3) the importance of intersection is analyzed by using entropy weight TOPSIS method, taking intersection connectivity, peak traffic flow and lane occupancy as evaluation indexes. The concrete steps of the importance evaluation of intersections are given, and the selection of key nodes in urban regional road network is realized. (4) considering only the correlation degree of two intersections in most traffic control subareas, the importance of key intersections is neglected. This paper presents a dynamic division method based on traffic control sub-area of key intersection. It uses key intersection as the starting point to traverse around intersection and use fuzzy inference to obtain the correlation degree between two adjacent intersections. The division of the control sub-area is determined by calculating the period difference between the coordinated intersection and the key intersection, and the proposed subarea division method is analyzed and studied with an example.
【学位授予单位】：兰州交通大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：U491.54
[Abstract]:With the rapid increase of motor vehicles, the problem of urban traffic congestion is becoming more and more serious, which restricts the development of economy and society, and brings great invariance to the travel of residents, so it becomes more and more important to solve the urban traffic congestion. Improving the control efficiency of traffic signals has become an effective method to solve traffic congestion. In view of the deficiency of the current signal control method and the control sub-area division method of intersection, combining with the intelligent control theory, the selection of signal control scheme and the importance evaluation of intersection are discussed in this paper. The dynamic division of the control sub-area is discussed and analyzed. The specific research contents are as follows: (1) the traffic signal control theory and fuzzy control theory are introduced systematically. The single stage fuzzy controller and two stage fuzzy controller are used in the intersection signal control. The control effect of the two controllers under different traffic flow is studied, and the selection of the input variables of the controller is analyzed. The configuration of controller structure under different traffic flow is determined. (2) aiming at the problem of single control model in single intersection signal control, it can not adapt to traffic flow change well. This paper presents a combined control model which combines single-stage fuzzy control and two-level fuzzy control, and realizes the switching of the two control strategies using SOM neural network. Aiming at the deficiency of manual setting of control rules and membership function of two-stage fuzzy controller, the parameters of two-stage fuzzy controller are optimized by chaos genetic algorithm. At the same time, in order to ensure the real-time control, combined with sliding time window, according to the real-time data acquisition, the fast identification of traffic state and the on-line optimization of two-stage fuzzy controller parameters are realized. Through the example simulation, the combined optimal control model is verified. (3) the importance of intersection is analyzed by using entropy weight TOPSIS method, taking intersection connectivity, peak traffic flow and lane occupancy as evaluation indexes. The concrete steps of the importance evaluation of intersections are given, and the selection of key nodes in urban regional road network is realized. (4) considering only the correlation degree of two intersections in most traffic control subareas, the importance of key intersections is neglected. This paper presents a dynamic division method based on traffic control sub-area of key intersection. It uses key intersection as the starting point to traverse around intersection and use fuzzy inference to obtain the correlation degree between two adjacent intersections. The division of the control sub-area is determined by calculating the period difference between the coordinated intersection and the key intersection, and the proposed subarea division method is analyzed and studied with an example.
【学位授予单位】：兰州交通大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：U491.54
【参考文献】
相关期刊论文 前10条
1 程海鹏;汤自安;汤e，
本文编号：2124736
本文链接：https://www.wllwen.com/kejilunwen/daoluqiaoliang/2124736.html
