Research on Delay Issues in the Active Control of Pantographs Based on Neural Networks

In this paper, to reduce the time delay in active control, an improved pantograph backstepping control scheme based on gated recurrent unit(GRU) was proposed. Firstly, a control-oriented nonlinear time-varying pantograph model was established. Secondly, the backstep controller was designed by using the Lyapunov stability. Finally, the contact force and pantograph state data are predicted by the GRU. The prediction backstep control based on the GRU predictor can significantly ensure the accuracy of prediction and real-time control, and then suppress the vibration of the pantograph system.