Single Neuron PID Based ESO for PMSM Control

A single neuron PID (SNPID) and extended state observer (ESO) is proposed, to address PID parameter optimization and disturbance compensation for PMSM systems. Firstly, it is difficult to determine the optimal values of the PID, a single neuron neural network is proposed to optimize the proportional, integral, and differential parameters, respectively. Secondly, in order to weaken the disturbance of the PMSM system, the ESO is proposed. Then, a composite control method with SN-PID and ESO (SN-PID-ESO) is used to improve the control quality. Finally, transient and steady state performances are verified by comparative simulations.