Some Issues On Iterative Learning Control Based High Speed Train Operation Control

This thesis focuses on some issues of iterative learning control based high speed train operation control. The repetitive nature of the high speed train allows iterative learning control to learn from previous runs and modify the control input for improved tracking performance. The main contributions of this work are summarized as follows:1. The focusing problem for train speed tracking, an adaptive iterative learning con-troller which contains iteration-varying compensation part is proposed. By fully using the relationship between the air resistance coefficient and the temperature, the iteration-varying air resistance coefficient is divided into iteration-invariant part and iteration-varying part. Iteration-invariant part is estimated using the adaptive updating law in iteration domain while iteration-varying part is compensated using real-time measured temperature information. The effectiveness of the proposed method is verified using a numerical example.2. Constraint based adaptive iterative learning controller which subject to a speed constraint during the operation of high speed train is proposed. In order to meet the se-curity requirement that the real-time operation speed can’t exceed the maximum speed limit in the corresponding railway section, barrier lyapunov function scheme is intro-duced into the control system. Thus the safety of train operation system is enhanced. Furthermore, data measured by the temperature sensors is used to eliminate the influence of the air resistance coefficient under different weather conditions. Finally, the simulation examples verify the effectiveness of the proposed algorithm.3. A robust adaptive iterative learning control based automatic train operation is proposed to address a high speed train tracking problem with consideration of iteration-varying operation condition and measurement noise. According to the relationship be-tween the air resistance coefficient and the wind pressure, the influence of the non-repetitive factor is eliminated by using the measured wind pressure. To alleviate the effect of measurement noise, a saturation function scheme is introduced into the pro-posed method, and the chattering is suppressed greatly. The boundedness of the system tracking error can be obtained if the measurement noise is boundedness. The Effective-ness of the proposed method is verified by rigorous mathematical analysis and numerical simulation.4. For multiple-point tracking of the high speed train, a norm optimal point-to-point iterative learning controller is designed based on a linearized model of the high speed train. Only using the fixed points’s information, thus lower computational complexity and less memory requirement is used comparing the traditional norm optimal iterative learning controller. In order to make the research close to the reality, the robustness of the proposed method under conditions of model uncertainty and wind gusts is analyzed. Both rigorous mathematical analysis and detailed simulation results confirm the correctness and effectiveness of the proposed method.5. An adaptive iterative learning controller for the coordination of multiple high-speed trains’ movement is proposed. The motion of an ordered set of high-speed trains running on a railway line is modeled by a multi-agent system. In this proposed control strategy, only communicates information with its neighboring trains is used to adjust itself. By utilizing the repetitiveness nature of the train operation, the unknown time-varying parameters are effectively estimated, and each train can track its own desired speed curve accurately. By introducing the artificial potential function, the headway dis-tances between any two adjacent trains are stabilized in a safety range. And the collision between the adjacent trains is avoided. Numerical examples are given to illustrate the effectiveness of the proposed methods.