A LSTM-NN-based Reference Trajectory Compensation Algorithm for the Maglev Planar Motor Control System

Maglev planar motor (MLPM) is a new kind of 6 degree-of-freedom (6-DOF) motor, which has attracted extensive attention, because it is very promising in precision and ultra-precision manufacturing, especially for semiconductor equipment. However, there is always a model mismatch between the traditional physics-based deduced (TPD) model and the real model of the motor control system, which leads to performance deterioration in designed control strategies. In the paper, a long short-term memory neural network (LSTM-NN) data-driven model is developed to reflect the dynamic characteristics of the MLPM control system accurately, and then, an LSTM-NN-based reference trajectory compensation algorithm is proposed to improve the tracking performance of the system. The corresponding control experiments are carried out and the results show that the developed data-driven model is able to accurately predict the position of the system, validating the accuracy of the model, and the trajectory tracking error is greatly decreased by applying the proposed LSTM-NN-based compensation algorithm. The effectiveness of the proposed algorithm is verified.