A parameters analysis of sine cosine algorithm on travelling salesman problem

Abstract

Sine Cosine Algorithm (SCA) is a fairly new algorithm developed in 2016 by Mirjalili, likewise Black Hole Algorithm (BHA), Whale Optimization Algorithm (WOA), Artificial Atom Algorithm (A3) and Physarum-Energy Optimization Algorithm (PEO) proposed in 2013, 2016, 2018 and 2019, respectively. Due to new ideas in SCA, a few publications have been published on SCA. SCA was applied to continuous and discrete optimization problems. Besides, there exist remarkable implementations of SCA in the field of engineering, science, and technology. In this work, a parameter analysis of SCA has been done on a classical TSP (Berlin52-CTSP) and randomly generated TSP (RTSP). To do parameter analysis, major parameters have been changed gradually. For classical TSP, symmetric data has been taken from TSPLIB (TSP Library in net). The results are given as best, mean, worst solutions, std. deviation and CPU time for CTSP and RTSP. Besides, figures and tables demonstrate the effect of parameters for solving TSP. After adequate experimentation, based on trial-and-error methodology, optimal parameters and best solutions have been found. As a result, the findings indicate that major parameters of SCA influence the performance of that algorithm significantly.