Neural-network-based self-tuning PI controller for Permanent Magnet Synchronous Motor

In the servo motor drive system applications, the variation of load inertia will degrade drive performance severely. Good dynamic and static performance of servo system requires controlling inertia robustly. In this paper, in order to get the moment of rotational inertia, online identification methods based on model reference adaptive identification (MRAI) were developed. Then a well-trained neural network supplies the PI controller with suitable gain according to each operating condition pair detected. To demonstrate the advantages of the proposed neural-network based on self-tuning PI control technique, the simulations was executed in this research. Result of simulations show that the newly developed dynamic PI approach outperforms the fixed PI scheme in rise time, precise track of angular velocity, and robustness when the inertia varied.