Quasi Oppositional—Manta Ray Foraging Optimization and Its Application to PID Control of a Pendulum System

Abstract

This paper presents an improved version of Manta Ray Foraging Optimization (MRFO). MRFO is relatively a single objective optimization algorithm. It was inspired from the behavior of a cartilaginous fish called Manta Ray. Manta Ray applies three strategies in searching foods which are chain, cyclone and somersault foraging. From the study, MRFO is a relatively new developed algorithm and has low convergence rate. However, MRFO has potential to be improved in that aspect. In the meanwhile, Opposition-based Learning (OBL) is a well-known technique in increasing the convergence rate. Therefore, a type of OBL namely Quasi Oppositional-based Learning will be adopted into MRFO in order to increase the possibility of finding the solution by considering the opposite individual location of fitness. This version of MRFO is called as Oppositional-based MRFO (OMRFO). Further, OMRFO was performed on several benchmark function. A statistical non-parametric Wilcoxon Test was conducted to analyze the accuracy of MRFO and OMRFO. Furthermore, the proposed algorithm was applied to an inverted pendulum system. Result from shows that performance of OMRFO is significantly outperformed MRFO after tested in the benchmark functions.

Manta Ray Foraging Optimization Algorithm: Modifications and Applications

“…Abdul Razak et al [29] offered a quasi-oppositional MRFO for proportional-integral-derivative (PID) control of a pendulum system. The study incorporated a quasi-based OBL to improve results by considering the opposite agent fitness positions.…”