Dual-redundancy PMSM servo system: Using single neuron PID controller

With the development of More-Electric/All-Electric aircraft technique, dual redundancy or more redundancy becomes a valid method for enhancing the reliability and security of motors in aircrafts. In this paper, based on the dynamic mathematic model of a 3.6kw dual-redundancy permanent magnet synchronous motor (DRPMSM), a single neuron PID controller is proposed in place of the conventional linear PID controller, thus improving the adaptability and robustness. The modified weight learning algorithm is obtained by combining the non-supervisory Hebb learning rule with the supervisory Delta learning rule to adjust the controller parameters on-line. By using this controller, the dynamic characteristics of the DRPMSM servo system was modelled and simulated by MATLAB/Simulink. The results showed that the single neuron PID controller could change its parameters automatically so enhanced dynamic performance of DRPMSM servo system and reduced torque fluctuation comparing with the system adopting conventional PID controller.