Performance Analysis of SMDO Method with Benchmark Functions with Matlab Toolbox

Abstract

Öz: SMDO method is a set and trial based optimization algorithm that is developed for onlinefine-tuning of controller parameters. SMDO method is implemented for several controller tuningapplications. It can search parameter space with random backward and forward steps of each parameter.This property reduces risk of testing unstable control system configurations in controller design and thusmakes the SMDO method more suitable for online parameter tuning of experimental systems. However,performance of SMDO has not been evaluated previously for benchmark functions in comparison withother well known heuristic optimization methods. This study aims to compare performances of ArtificialBee Colony (ABC), Cuckoo Search Optimization (CK), Particle Swarm Optimization (PSO) andStochastic Multi-parameters Divergence Optimization (SMDO) methods for benchmark functions.Therefore, a benchmark tests program that is a user-friendly MATLAB GUI is introduced for user. Thisprogram can be downloaded from https://www.mathworks.com/matlabcentral/fileexchange/75043-smdo-method-with-benchmark-functions.