Adaptive Robust Control Of Train Operation

Over the last decade, the explosion of High-Speed Rail and Urban Mass Transit provides a strong basis for the fast and sustainable growth of Chinese Econo-my. In the next period of time, the target of improving the operational efficiency and performance becomes the developing tendency of Railway Transit Equipment. Since the performance of Automatic Train Operation affects the operational efficiency and energy consumption of the whole railway system, it becomes a hot research topic in engineering field of railway transit.In this dissertation, the single-point and multi-point dynamic model of the train are built respectively, in presence of parameterized uncertainty, time-varying property, input saturation nonlinearity, and uncertain nonlinear coupling forces between adjacent vehi-cles of the train. On this basis, the adaptive and robust control theory are explored to achieve the tracking objective of ATO. Finally, a kind of high-performance controller is designed for the general complex multi-point train model.The main contents and contributions of this dissertation are listed as follows:Firstly, with consideration of the time-varying parameterized uncertainties, the single-point dynamic model of the train is constructed, and its uncertainties and time-varying properties are analyzed qualitatively. Then, a backstepping adaptive controller is designed to compensate the parameterized uncertainties, and the stability analysis of the closed-loop system is also given. On this basis, according to the uncertainty bound of train dynamics, a deterministic robust control law is designed to reject the time-varying properties of model parameters, and the theoretical analysis that train tracking errors are converging to the desired accuracy range with designed convergence rate, is also given.Secondly, with consideration of jumping parameters, an Immersion and Invariance based estimator is designed for the unknown parameters. Not only the I-type adaptive law, but a P-type tuning function is added into adaptive estimation, so the proposed estimator can provide a larger degree freedom to improve convergence property of adaptive esti-mation. With the aid of certain output error transformation, the prescribed performance is introduced into the original system, then a performance-guaranteed adaptive robust controller is designed. The proposed controller eliminates the influence of time-varying parameters and unknown external disturbances, and theoretical analysis also shows that train tracking error would guarantee the prescribed performance for all time.Additionally, with consideration of train traction and braking limitations, a contrac-tion based adaptive control strategy is proposed, and a desired compensation based update law is designed for the unknown parameters to bound the adaptive compensation control. Then, according to the residual power of the train, a saturated adaptive robust controller is synthesized, and it renders train tracking errors converging to the desired accuracy range within a finite time. Besides the global stability analysis of the closed-loop system is also given.Finally, with consideration of the uncertain nonlinear in-train forces and the dynam-ic differences between vehicles at different locations in one train, the multi-point time-varying nonlinear model of the power-distributed train is built. Based on the foregoing achievements, an adaptive robust controller is designed for the power-distributed train-s, and the theoretical analysis of system stability and train tracking errors exponentially converging to the desired accuracy range is also given. The simulation results verified that not only train tracking performance is improved, but the in-train forces and energy consumption are also reduced during train operation.