Özet:

This article provides a modified version of the atomic search optimization (ASO) algorithm using opposition-based learning (OBL) to improve search space discovery. OBL is a machine learning strategy that is widely used to improve the performance of intuitive-up algorithms. The counter-based ASO (OBASO) algorithm, presented as a new design method, was first recommended in finding the optimal values of the proportional-integral-functional plus second-degree derivative (PIDD2) controller parameters in the automatic voltage regulator (AVR) system. In the design problem, a new target function, which includes the integration of the time-growing square error (ITSE) and the exaggeration together, optimized with the recommended OBASO algorithm, found the best values of the PIDD2 controller parameters. The performance of the recommended OBASO-adjusted PIDD2 (OBASO-PIDD2) controller is compared with the classic ASO-adjusted PIDD2 (ASO-PIDD2) controller, as well as the PID, Cutting Degree PID (FOPID) and PIDD2 controller, adjusted by modern intuitive-upper algorithms. Comparative temporary state and frequency response analyses have been carried out to assess the stability of the proposed approach. Furthermore, the rigidity of the recommended approach has been tested taking into account possible changes in the AVR parameters. The comprehensive simulation results and comparisons with other existing controls show that the recommended OBASO-PIDD2 controller with a new target function has excellent control performance and can significantly increase the system rugby due to model uncertainty.

Anahtar Kelimeler: