关键词: 列车自动运行, 迭代学习辨识, 迭代学习控制, 学习律

Abstract: Aiming for the difficulty to establish accurate dynamic model for train control system, and combining a large number of duplicate information contained in the train operation, the unknown parameters in the train dynamics model are identified by using Iterative Learning Control（ILC） algorithm and a control algorithm based on ILC is proposed to control the train. The core of the algorithm is the use of historical data to generate a new input to control train. The simulation results show that after a certain number of iterations, parameters identification value remains stable and the train can strictly follow the target curve to run and ensure the train can travel with high-precision, high-steady and high-security.

Key words: automatic train operation, iterative learning identification, iterative learning control, learning law