Decentralized Control on High-Speed Trains Based on State Observation

It is well known that the dynamic model of high-speed trains are difficult to be accurately modeled, due to the air resistance shows nonlinear characteristics, and other resistances also be obtained by some empirical formulas. The above resistances are regarded as a real-time disturbance for train. Hence it has a significant adverse impact on train operation. Here, a cooperative controller based on a expansion state observation is proposed to estimate these resistances. The proposed control allows each car to be controlled asynchronously according to the railroad conditions in order to achieve high tracking performance of the speed profile and minimization of internal forces, which will greatly improve the punctuality and safety of high-speed trains. In addition, the observer is also designed based on a reduced-order system to reduce the design parameters. Train and coupler lengths are also considered to fully analyze the train dynamics associated with interaction forces as the train passes through complex road conditions.