数据驱动控制方法在交通控制中的应用研究

发布时间：2018-01-15 01:06

本文关键词：数据驱动控制方法在交通控制中的应用研究　出处：《北京交通大学》2014年博士论文　论文类型：学位论文

【摘要】：随着机动车保有量的迅速增长,交通拥堵现象越来越司空见惯。如何合理利用现有交通设施,通过提高管理和控制水平,使交通系统充分发挥其能力,提高道路运行效率,成为解决交通问题的有效途径。本文研究了将数据驱动控制方法应用于过饱和单交叉口排队长度均衡控制、干线相关占有率均衡控制、快速路出入口匝道控制等若干问题。论文的主要研究内容和创新点总结如下：第一、基于存储转发建模方法,将迭代学习控制应用于过饱和单交叉口的排队长度均衡控制中,使交叉口各相位的排队长度趋于一致,考虑了交通需求满足严格重复、非严格重复等情况,给出了相应的收敛性分析,并通过PARAMICS仿真验证了有效性。第二、为了减少在线计算量,给出了基于多参数规划的单交叉口均衡控制方法,定义了排队长度均衡的性能指标,离线得到了信号配时控制律的显式表达形式,在线根据排队长度参数向量所在的凸区域便可求得对应的信号配时策略,并通过PARAMICS验证了其有效性。第三、将匝道控制的基本思想应用于过饱和主干线控制中,给出了相关占有率临界同步和临界相关占有率的定义,通过动态调整上下游交叉口的信号配时策略,使各支路的相关占有率趋于均衡,避免产生绿灯时间损失、溢流、死锁等现象,主干线系统输出达到最大,并通过PARAMICS验证了策略的有效性。第四、基于快速路实际运行产生的检测数据,将迭代反馈整定、虚拟参考反馈整定等若干数据驱动控制方法应用于快速路ALINEA控制器参数整定,通过PARAMICS验证了有效性。将自抗扰控制方法应用于快速路入口匝道控制中,通过PARAMICS验证了有效性。针对城市快速路出口匝道车辆无法正常、快速地驶离主路的现象,给出了主路出口匝道与辅路下游交叉口协调控制算法,仿真验证了该策略可提高快速路服务水平。
[Abstract]:With the rapid growth of vehicle ownership, traffic congestion is becoming more and more common. How to make rational use of existing traffic facilities, through improving the level of management and control, make the traffic system give full play to its ability. Improving the efficiency of road operation has become an effective way to solve traffic problems. In this paper, the data-driven control method is applied to the queue length equalization control of supersaturated single intersection and the trunk correlation occupation rate equalization control. The main contents and innovations of this paper are summarized as follows: First, based on the storage and forwarding modeling method, the iterative learning control is applied to the queue length equalization control of over-saturated single intersection, which makes the queue length of each phase of intersection tend to be consistent. Considering the case of strict repetition and non-strict repetition of traffic demand, the convergence analysis is given, and the validity is verified by PARAMICS simulation. Secondly, in order to reduce the amount of on-line calculation, the equalization control method of single intersection based on multi-parameter programming is presented, and the performance index of queue length equalization is defined. The explicit expression of the signal timing control law is obtained offline, and the corresponding signal timing strategy can be obtained online according to the convex region of the queue length parameter vector. Its effectiveness is verified by PARAMICS. Thirdly, the basic idea of ramp control is applied to the supersaturated trunk line control, and the definitions of critical synchronization and critical correlation share are given. By dynamically adjusting the signal timing strategy of the upstream and downstream intersections, the relative share of each branch tends to be balanced, avoiding the loss of green time, overflow, deadlock and other phenomena, the main line system output reaches the maximum. The effectiveness of the strategy is verified by PARAMICS. In 4th, based on the detection data generated by the actual operation of the expressway, several data-driven control methods, such as iterative feedback tuning, virtual reference feedback tuning and so on, are applied to the parameter tuning of the ALINEA controller of the expressway. The effectiveness of ADRC is verified by PARAMICS. The ADRC method is applied to the on-ramp control of expressway. The validity is verified by PARAMICS. In view of the phenomenon that the vehicle on the exit ramp of the urban expressway can not be normal and leave the main road quickly, the coordinated control algorithm of the exit ramp of the main road and the downstream intersection of the auxiliary road is presented. The simulation results show that the strategy can improve the service level of expressway.
【学位授予单位】：北京交通大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：U491.54

【参考文献】