A Hybrid DE - HS Algorithm with Randomized Parameters

Abstract

The evolutionary algorithms and their hybrid methods are quite efficient and accurate in terms of solution quality of optimization. In this study, a new hybrid algorithm is generated by merging Differential Evolution (DE) and Harmony Search Optimization (HS) algorithms which is called DES. The core steps of the algorithms are used without any modifications, but the main control parameters which directly affect the performance are randomized. The experimental study is done by comparing DE, HS and their hybrid method DES. According to the results, it is found that DES algorithm has improved the performances of original algorithms for the selected test problems.