Automatic Stop Control Of Electric Multiple Units

In recent decade, the rail transit system, which is represented by urban rail transit and high-speed railway, has been established in China. The scale of China’s railway network has reached the international leading level. With the expanding of railway network, the design of high-level performance train control system has become the premise of improving the railway operation efficiency, and the stop accuracy is an important index to measure the control performance of the train control system. Consequently, the automatic stop control is a significant subject in the field of the rail transit control nowadays.The automatic stop control of the rail transit system is studied in this dissertation. Physical properties of every link in the brake system and dynamic characteristics of the train are analyzed first. To deal with nonlinearity, time-delay in the brake system and uncertain resistance, several single-particle models are built, and the corresponding stop controllers are designed. The stability proofs are also proposed. Then a multi-particle dynamic model is constructed, considering characteristics of the coupling between carriages and inhomogeneous distribution of the resistance. Meanwhile, a decentralized automatic stop controller and the relative stability proof are presented.The main contents and contributions of this dissertation are listed as follows.First of all, based on dynamic characteristics of every link in the brake system, a linear brake model with time delay process is built. A brake deceleration controller is designed to restrain the brake impact effectively. Considering the accurate measurement of brake deceleration and the nonlinearity of the foundation brake rigging, a second order nonlinear model with input delay is constructed. The state prediction expression of the nonlinear system with input delay is built by functional operators. Based on the prediction expression, a Backstepping brake controller is designed. The theoretical proof of the convergent behavior of tracking errors is also presented.Secondly, dynamic characteristics of the mechanical brake system are analyzed. A third order linear dynamic model of the stop system, with the assumption of constant brake shoe coefficient, is built. To track the ideal stop curve swiftly, a model reference adaptive controller is designed based on PIQ and Backstepping technique. The theoretical proof of the stability of the closed-loop stop system and convergent behavior of velocity tracking error is also presented. Then the robust adaptive laws, which are based on projection method, are designed to handle the uncertain resistance. Meanwhile, the method of variable error coefficient is proposed to deal with the temporary impact of the control signal. The theoretical proofs of the system stability are given respectively.Thirdly, considering the time delay of the brake system, the variation of the brake shoe friction coefficient along with the running speed of the train and the disturbances in the brake system, a third order nonlinear dynamic model of the stop system with input delay and disturbances is constructed. Using functional operators, the state prediction relationship of the nonlinear system with input delay is established on condition of bounded disturbances. Based on this relationship, a Backstepping robust automatic stop controller is designed. The proposed controller can decrease the computational complexity of the Backstepping technique. Finally, the theoretical proof that the stop system is Input-to-State-Stable with respect to the bounded disturbances is presented.Finally, considering the longitudinal dynamics of the train, inhomogeneous distribution of the resistance, and the dynamic characteristics and switching of pneumatic-electric blending brake system, a multi-particle nonlinear model with unmodeled dynamics is built based on large-scale system theory. Then the nonlinear model is transformed into a T-S fuzzy model according to the running speed of the train. To approximate the upper bound of unknown interconnected terms and unmodeled dynamics, a robust adaptive fuzzy system is constructed. Based on this fuzzy system, a distributed sliding mode automatic stop controller is designed. The theoretical proof of the convergent behavior of the tracking errors is presented at last.