Abstract:

      In view of the difficulty in dealing with nonlinear disturbance moments such as friction in the flight simulator’s servo system, a compound control method using a brain emotional learning (BEL) model is proposed. Based on the traditional PI feedback controller, a BEL model-based identifier and a forward compensation controller of the same structure are designed. The node weights of the BEL model are obtained through inverse model identification of the system. Simulation results of an ideal simulator model and experimental results of a real simulator servo system show that the BEL model has good learning ability and meets the real-time control requirement. Moreover, the compound control strategy based on the BEL model can effectively reduce the influence of friction moment and improve tracking performance of the simulator.

Key words: flight simulator , brain emotional learning , system identification ,   compensation control