A High-Speed ​​Train Operation Control Method Based on Adaptive Neural Networks

Abstract

This invention discloses a high-speed train operation control method based on an adaptive neural network. The method includes: establishing a longitudinal dynamic model of the high-speed train; observing the train's speed and acceleration states in real time using a high-order sliding mode observer; and combining this with a neural network to identify unknown system dynamics, thereby improving dynamic approximation accuracy. An adaptive output feedback controller is then used to achieve precise tracking control of the train's displacement and speed by combining the observed train state with the approximated unknown dynamics. This controller, by introducing a leakage term, suppresses the accumulation of parameter estimation errors and enhances the train's adaptability to uncertainties. The control method proposed in this invention can effectively improve the stability and trajectory tracking accuracy of train operation and ensure the safe and reliable operation of the train under unknown dynamic disturbances.

Images