An Analysis of the Parameter Modifications in Varieties of Harmony Search Algorithm

Abstract

A Harmony Search (HS) algorithm is a population based-meta-heuristics approach that is superior in solving diversified and large scale optimization problems. Several studies have pointed out that Harmony Search is an efficient and flexible tool to resolve optimization problems in diverse areas of construction, engineering, robotics, telecommunication, health and energy. Considering its increasing usage in diverse areas, this paper aims to present the historical development of HS, highlighting on its different features, modifications for improvements and limitations. Based on the description of the fundamental concept of HS, recent variations of the extended HS were analyzed focusing on its algorithm’s theory as well as its fundamental and primary concepts. It was found that the enhancements made on the extended HS are mainly on the modification of the parameters, such as the harmony memory consideration rate (HMCR), pitch adjusting rate (PAR) and distance bandwidth (BW). This analysis provides a useful motivation for researchers interested to improve the achievement of the standard HS algorithm and enhance the solution convergence rate and flexibility.