Abstract
Starting from the classic FOC (Field Oriented Control) structure of a PMSM (Permanent Magnet Synchronous Motor), this article presents the comparative performance of the control system when the tuning of the PI-type speed controller is performed both in the conventional way (using the Ziegler-Nichols method supplemented by the trial and error method), and using three computational intelligence specific algorithms, namely: Genetic Algorithm (GA), Simulated Annealing (SA) algorithm, Reinforcement Learning (RL) - Twin Delayed Deep Deterministic Policy Gradient (TD3) agent algorithm. The superiority of the results of the use of computational intelligence algorithms for tuning the PI-type speed controller is demonstrated by numerical simulations performed in the Matlab/Simulink development environment.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
