【作者】 刘辉；

【导师】 孙棣华；

【作者基本信息】 重庆大学 ， 控制理论与控制工程， 2017， 博士

【摘要】 在近信号控制区域,连续车流在动态变化的信号灯影响下呈现出周期性的驶停,并极易造成燃油浪费,甚至引起追尾碰撞等交通事故。因此,掌握车辆在近信号控制区行驶的内在规律,并探索相应的控制机制,对于提高通行效率和改善交通安全具有重要意义。现有方法主要从辅助驾驶和自动驾驶两个角度展开,分别研究了个体车辆以能耗或效率为目标的速度引导算法和面向多车的集中式协同控制策略。事实上,近信号控制区车辆行驶问题,究其本质是车车和车路协同问题。因此,以协同行驶为切入点将有助于更深入地揭示车辆在近信号控制区行驶的机制。同时,随着信息化技术在近信号控制区的广泛应用使得车辆在这一区域的行驶过程具有典型的信息物理融合特征。为此,以信息系统与物理系统融合为突破口,充分发挥交通信息系统的优势实现交通物理系统的协同,这一途径可望成为新环境下解决近信号控制区车辆协同行驶问题的有效方法。为此,本文从交通信息物理融合视角,以车车协同行驶一致性特征为突破点,围绕近信号控制区车辆协同行驶控制方法展开研究。首先提出了车辆协同行驶一致性的定性和定量描述方法。其次,针对交通信息物理融合环境下部分车辆可控情形,研究了考虑行驶一致性的车车协同控制方法。再次,针对信号灯约束下的车辆协同行驶问题,研究了单车行驶的优化控制方法和考虑多后车行驶一致性的优化控制方法。最后从交通信息物理融合视角研究了面向车路协同头车的近信号控制区跟随车协同行驶控制策略,并探索了集中式控制与分布式控制相结合的协同行驶系统的一致性分析方法。本文完成的主要研究工作包括:(1)从协同行驶系统的本质特征出发,提出了行驶一致性的定性和定量描述方法。车车协同问题是交通信息物理系统(Transportation Cyber Physical Systems,T-CPS)的研究热点之一,一直以来广受学者关注。现有研究聚焦于协同策略的制定和协同系统的搭建等问题,而协同行驶性能是协同策略和协同系统的核心和基础,因此建立合适的协同行驶评价方法至关重要。本文首先根据扰动在协同行驶系统中的传播规律,研究了车辆协同行驶实现一致的定性条件。进一步针对协同行驶系统的动态特征,基于车速差和车头距等建立了协同行驶一致性的定量描述方法。最后通过仿真实验分析了所提行驶一致性与宏观交通流之间的关系。(2)考虑路段中部分车辆可控情形,提出了交通信息物理融合环境下考虑行驶一致性的协同行驶优化控制方法。针对车辆控制难度大和成本高的问题,本文首先建立了基于动态限速控制的车辆控制架构,借助交通信息物理融合环境下邻近车辆的状态信息和(1)中的行驶一致性度量方法构建了协同行驶优化目标,并利用车辆对动态限速的响应模型构建了基于非线性模型预测控制的优化控制方法。最后分析了路段部分车辆可控和路网中部分路段可引导情形下的协同行驶系统性能,仿真结果证明了所提方法的有效性。(3)考虑动态变化的信号灯对协同行驶的约束,建立了近信号控制区车车-车路协同行驶优化控制方法。在工作(2)的研究基础上,针对存在信号灯约束下的协同行驶问题,首先建立了近信号控制区车辆协同行驶模型,并分析了信号灯信息对车辆协同行驶系统的影响,分析发现信号灯切换过程对行驶一致性影响较大。其次,从交通信息物理融合视角,以单车的行驶一致性为目标建立了优化控制方法。进一步,考虑部分车辆可控情形,研究了考虑多后车行驶一致性的协同控制方法。最后借助仿真实验证明了本文方法的有效性和优越性。(4)面向车路协同头车,从交通信息物理融合视角提出了近信号控制区跟随车协同控制方法。在工作(3)的研究基础上,将近信号控制区车辆协同行驶系统构建为分布式控制与集中控制相结合的控制系统。领头车采用(3)的优化控制方法,针对跟随车的协同控制问题,首先借助交通信息物理系统的优势,建立了考虑邻车控制信息的跟随车一致性控制策略。并分析了通信时延对协同行驶系统实现一致性的影响。此外,针对通信网络结构存在动态变化问题,应用Lyapunov稳定性理论和线性矩阵不等式技术研究了协同行驶系统实现一致性的条件,仿真结果证明了所提方法的有效性和理论分析的正确性。综上所述,本文深入研究了交通信息物理融合环境下近信号控制区车辆协同行驶一致性控制方法,系统的讨论了行驶一致性的描述方法及相关的优化控制方法等问题,理论分析和仿真实验均验证了上述工作的有效性。研究成果可望为信号交叉口协同系统的工程实践提供新方法,还可为交通信息物理系统的优化控制提供理论支撑。更多还原

【Abstract】 In the adjacent area of traffic signal control,under the influence of dynamic changes of the signals that the continuous traffic flow become a form with periodically stop and go,and it easy lead to the waste of oil and even the traffic accident such as rear-end et al.Therefore,it is necessary to improve the traffic efficiency and safety by the mastering the internal rule of vehicles driving in the adjacent area of traffic signal control and exploring the corresponding control mechanism.The existing methods mainly from the aspect of assistant driving system and automatic driving system,then the speed guidance algorithm for vehicle’s fuel consumption and time save and the centralized coordination control strategy for multi-vehicles are studied respectively.Actually,the vehicles’ driving process in the adjacent area of traffic signal control which is relate to cooperation of vehicle to vehicle and vehicle to infrastructure naturally.Therefore,it is helpful for revealing the mechanism of vehicle driving in the adjacent area of traffic signal control from the idea of cooperative driving.At the same time,with the widely use of the information technology in the adjacent area of traffic signal control that the process of vehicle driving in the adjacent area of traffic signal control can be characteristic as a typical cyber-physical system.Therefore,from the aspect of integrate the cyber system and physical system and take the full advantages of transportation information system to realize the cooperation of transportation physical systems,and it is expected to become an effective method to resolve cooperate driving problem in the adjacent area of traffic control for new environment.To this end,here we take the cooperative driving consensus as the breakthrough point from the perspective of transportation cyber-physical system to study the cooperative control method for vehicles driving in the adjacent area of traffic signal control.At first,the qualitative and quantitative description method for cooperative driving consensus is proposed.Secondly,under the condition of transportation cyber-physical that a cooperative control method by the consideration of driving consensus for part of the vehicles is studied.Thirdly,for the condition that vehicles constrained by traffic signal,the cooperative optimal control scheme for optimizing single vehicle and multi-vehicles’ driving consensus are studied.At last,the cooperative control method for vehicles with a cooperative vehicle to infrastructure leader from a transportation cyber-physical view,and the consensus analysis method for centralized and distributed cooperative driving system is studied.The main research work includes:(1)Based on the essential characteristic of the cooperative driving system,the qualitative and quantitative description method for driving consensus is proposed.The cooperative vehicle to vehicle is a hotspot for transportation cyber-physical system and it has been studied widely.Recent researches mainly focused on the cooperative method and the construct for cooperative system.However,the performance of the cooperative driving system is the core and foundation for the cooperative method and system,and establish a proper evaluation method for cooperative driving system is very important.In this thesis,a qualitative description method is studied based on the rules of disturbance propagation in cooperative driving system.Moreover,according to the dynamic characteristic of the cooperative driving system,a quantitative description method for driving consensus is established by the use of velocity difference and headway.At last,the relationship between driving consensus and traffic flow is analyzed through the simulations.(2)Considering the condition that part of the vehicles can be controlled in the traffic system,the driving consensus based cooperative control method is proposed from transportation cyber-physical system view.Due to the difficulty and cost of the vehicle control,a variable speed limit control based vehicle control framework is established,and the optimization objective is constructed by the use of neighbor vehicles’ information and the driving consensus which has been defined in(1),and a nonlinear model predictive control based optimizing control method is established according to the vehicle’s dynamic model which relate speed limit with vehicle’s performance.At last,the cooperative driving system’s performance is analyzed under the condition that part of the vehicles or vehicles in part of the road can be controlled.The simulation results demonstrated the effectiveness of the proposed cooperative control method.(3)Considering the constraints of dynamic traffic signal,the cooperative driving control method is established based on cooperative vehicle to vehicle and vehicle to infrastructure for vehicles driving in the adjacent area of traffic signal control.Based on the studied in(2),for the cooperative driving which constrained by dynamic traffic signal,the cooperative driving model is established for vehicles driving in the adjacent area of traffic signal control,and the driving consensus is analyzed based on the model,the result show that the process of traffic signals change affect the driving consensus obviously.Secondly,the optimizing control method for single vehicle is established by consideration of driving consensus from transportation cyber-physical system perspective.Moreover,aim at the condition that part of the vehicle can be controlled,the optimizing control method with the consideration of multi-vehicles’ driving consensus is studied.At last,simulation results are provided to demonstrate the effectiveness and superiorityof our method.(4)Based on the leader vehicle with cooperative vehicle to vehicle and infrastructure,the cooperative control method for following vehicle in the adjacent area of traffic signal control is proposed from transportation cyber-physical view.Based on the studied in(3),the cooperative driving system for vehicles in the adjacent area of traffic signal control is rebuilt in a form which integrate distributed and concentrated control.The leader vehicle is controlled by the use of the scheme in(3),and take advantage of the transportation cyber-physical system,the cooperative control for following vehicles is proposed by the use of neighbor vehicles’ states and control signals.Then,the communication time delays impact on consensus of the cooperative driving system is analyzed.Moreover,for the condition that the cooperative driving system with switch communication topology,by the use of Lyapunov theorem and linear matrix inequation that the condition for cooperative driving system achieve consensus is studied.The simulations demonstrate the validity and correctness of the theoretical analysis.In conclusion,the cooperative control method for vehicles driving in the adjacent area of traffic signal control by the consideration of driving consensus from transportation cyber-physical view is been studied deeply in this thesis.The description method for driving consensus and optimal control method with the consideration of driving consensus is been studied systematically.In addition,the theoretical analysis and simulation results provided to demonstrate the effective of the researches.The research results will expecte to provide new methods for the engineering practice of the cooperative driving system,and it can support the optimization control theory for transportation cyber-physical system.更多还原