In this work, a novel distributed fault-tolerant control strategy is proposed for cooperative operation of high-speed trains (HSTs) subject to unmodeled dynamics, time-varying external disturbances, input saturation and actuator partial failures. The cooperative control of HSTs is first transformed to the consensus problem of a multi-agent system. Then, the distributed control laws that are equipped with a command filter and an adaptive neural network are developed to achieve the consensus task among the HSTs with predefined displacement and speed profiles. Further, an auxiliary dynamical system is introduced to compensate for the influence of input saturation. Moreover, an adaptive failure compensation strategy is designed to deal with the unknown and time-varying actuator faults. The convergence of the proposed controller is analyzed rigorously by applying the Lyapunov theorem, and the effectiveness is demonstrated by numerical simulations.

The existed control methods of the multi-area load frequency control (LFC) system fail to decouple the total tie-line power flow. This defect can be addressed as a problem on how to effectively utilize the total tie-line power flow. To improve that defect, the energy and structure of multi-area LFC system have been carefully studied, such that, the energy goes through the systemic cascade parts and the systemic structure matrix is partially skew symmetry. Namely, the energy and structural properties of the multi-area LFC system are similar to the properties of Port-Hamiltonian (PH) system and cascade system. Inspired by the above properties, a new method based on the PH system and cascade system that is proposed to design some PID control laws for the multi-area LFC system successfully works out the aforementioned problem. Compared with the existed PID methods for the multi-area LFC system, the proposed method has two advantages, which are the decoupling of total tie-line power flow and the robust disturbance rejection. At last, simulations results demonstrate the validity and advantages of the proposed method.

This paper studies the distributed data-driven event-triggered model free adaptive iterative learning control (ETMFAILC) of multiple high-speed trains (MHSTs) under iteration-varying topologies, which breaks away from the dependence on the train dynamics. Firstly, the nonlinear MHSTs with unknown dynamics are converted into a linear model. Then, combining the proposed event-based triggering condition and the linear model, the ETMFAILC scheme under the fixed topology is designed. Next, theoretical analysis proves the bounded input bounded output (BIBO) stability of MHSTs. Finally, the study is extended to the switching topologies and the validity of the ETMFAILC is verified by a numerical example.

In whole operation procedure, trajectory tracking control for heavy haul train with the time-varying desired speed is investigated in this paper. Taking full advantage of the heavy haul train's facilities, a proportional-integral (PI) control method based on the real-time speed error and the real-time displacement is proposed to represent the dynamic model of heavy haul train as port-Hamiltonian (PH) system framework, which implements the interactive feature of the in-train force and the damping feature of the basic resistance force effectively. Due to the proposed method, the real-time speed of heavy haul train asymptotically tracks the time-varying desired speed and the real-time displacement is small in whole operation procedure. Since the basic resistance force is utilized to construct the nonnegative definite matrix and heavy haul train is represented as PH system, the proposed method is also effective for trajectory tracking control of heavy haul train subjected to the uncertain and time-varying basic resistances without parameter adaption or estimation. In short, a universal tracking control scheme is presented for heavy haul train subjected to the aforementioned different operative conditions. Finally, simulations confirm the validity and advantage of the proposed method.