Longitudinal Stability Analysis And Control Of Stochastic Vehicles Following System

Automated Highway System (AHS) is an uppermost subsystem of Intelligent Transport System (ITS). The system includes vehicle automatic navigation and control, traffic management automation and traffic accident treatment automation. It will significantly increase the throughput on the highways and enhance driving safety. A key factor in AHS deployment is to analyze the stability and control of the vehicle longitudinal following system for vehicle automated driving. In this dissertation, the random vehicles longitudinal dynamics model was set up based on the random factors, the stability of several types of nonlinear stochastic interconnected system was analyzed, also the control and other related content of two kinds of automated highway random vehicles longitudinal system were studied on the basis of the achieved stability criteria.First the main random factors ignored in the vehicle system modeling were analyzed, the "look ahead" and "look both ahead and behind" stochastic vehicles longitudinal dynamics model were established respectively by using ITO stochastic differential equations, and then on this basis, the "look ahead" and "look both ahead and behind" stochastic vehicles longitudinal dynamics model with time delay were established respectively by taking the time delay factors into account.The analysis was also made for the stability of infinite-dimensional stochastic interconnected systems secondly. With the deterministic box theory being generalized to the corresponding stochastic box theory, under the assumption that the excitations are parametric white noises, the stability of infinite-dimensional autonomy stochastic interconnected systems was studied by making use of the Ito stochastic differential equation and the vector Lyapunov function method, and then the sufficiency criterions for the stability of exponential mean square sting were obtained. In the meanwhile, by considering the time delay factors, the stability of the infinite-dimensional autonomy stochastic interconnected systems with time delay was studied and a sufficient condition was derived. Then the stability of the infinite-dimensional non-autonomy stochastic interconnected systems and the one with the time delay was analyzed, the stability criterions obtained as well.Next, the analysis focuses on the stability and control of the "look ahead" stochastic vehicles longitudinal following system in AHS. Based on the stochastic stability conditions of isolated subsystems, the sufficiency for the exponential string stability was achieved by making use of the Ito stochastic differential equation and the vector Lyapunov function method. Taking the criterion as theoretical basis, the controller of the "look ahead" stochastic vehicles longitudinal system was designed by adopting the sliding mode control method. Based on the random vehicles longitudinal dynamics model, the state equation of the vehicles longitudinal following system is established; the corresponding control law was designed by assuming that the selection of vehicle actuator or the brake force for the control variables is according to the displacement between the workshop communication receiving leader and the adjacent vehicles before it and the velocity information in the fleet; the stability of the control system was analyzed by applying the obtained criterion making certain the exponential group stability. Based on the "parameters are bounded" and "parameters varying slowly" two types, the stochastic vehicles longitudinal system with parameter uncertainties was studied and robust control strategies and parameters adaptive law were designed.After that, the stability and control of the "look both ahead and behind" stochastic vehicles longitudinal following system in AHS was analyzed. The string stability of the AHS vehicle longitudinal following system was studied by the use of the vector Lyapunov function method. The sufficiency criterion was obtained for the exponential string stability of the infinite-dimensional stochastic vehicles longitudinal following systems. The controller of the " look both ahead and behind " stochastic vehicles longitudinal following system was designed by adopting the sliding mode control method. The stability of the control system was analyzed by applying the obtained criterion. Based on the "parameters are bounded" and "parameters varying slowly" two types, the "look both ahead and behind" stochastic vehicles longitudinal following system with parameter uncertainties was studied and the corresponding control strategies were designed.Then, the stability and control of the "look ahead" stochastic vehicles longitudinal following systems with time delay in AHS was studied. By making use of the stochastic box theory and the vector Lyapunov function method, the stability of the "look ahead" stochastic vehicles longitudinal system with time delay was analyzed and a sufficient condition was derived. The state equation of the vehicles longitudinal following system was established according to the random vehicles longitudinal dynamics model with time delay; the corresponding control law was designed by adopting the sliding mode control method; the stability of the control system was analyzed by applying the obtained criterion insuring the exponential string stability of the stochastic vehicles longitudinal system with time delay; the time-delay stochastic vehicles longitudinal system with parameter uncertainties was studied and robust control strategies and adaptive algorithm to compensate for parametric variations was designed.After then, the analysis was made for the stability and control of the "look both ahead and behind" stochastic vehicles longitudinal following system with time delay in AHS. The model of the infinite-dimensional stochastic vehicles longitudinal following system with time delay was established by making use of the Ito stochastic differential equation. By applying the stochastic box theory and the vector Lyapunov function method, the stability of the "look both ahead and behind" stochastic vehicles longitudinal following system with time delay was analyzed. The sufficiency criterion was obtained for the exponential string stability of the infinite-dimensional stochastic vehicles longitudinal following system, based on which theoretically, the corresponding state equation was obtained by combining with the random vehicles longitudinal dynamics model. The controller of the "look both ahead and behind" stochastic vehicles longitudinal following system was designed by adopting the sliding mode control method and the stability was analyzed. Based on the "parameters are bounded" and "parameters varying slowly" two types, the "look both ahead and behind" stochastic vehicles longitudinal following system with parameter uncertainties was studied, meanwhile, robust control strategies and adaptive algorithm to compensate for parametric variations was designed.Finally, the simulation was done by applying Matlab toolbox. Simulation results indicate that two types of control laws for the stochastic vehicles longitudinal following system designed in the dissertation were efficient and plausible.