High-Precision Beam Steering Mechanism Control Based on Model Compensation and Neural Network

In this paper, a control method is proposed based on model compensation and neural network to improve the tracking performance of a beam steering mechanism (BSM). The dynamic model of the fast-steering mirror (FSM) is first established. Then, the extended state observer (ESO) can be compensated through the known model information to improve the ability of ESO to observe interference. Besides, the approach adjusts the control parameters through the result of system identification by diagonal recurrent neural network (DRNN) to simplify the tuning process. Finally, numerical simulations are conducted to validate the effectiveness of the proposed control method.