Enhanced harmony search with dual strategies and adaptive parameters

Abstract

Harmony search (HS) is an emerging approach for global optimization. However, HS often demonstrates insufficient convergence due to the weak exploitation capability of its search strategy. Concerning this weakness, an enhanced HS with dual strategies and adaptive parameters (DSAHS) is proposed. In its search process, DSAHS conducts the best harmony-guided and random harmony-guided search strategies to balance the exploration and exploitation capabilities. Moreover, DSAHS adaptively adjusts its major parameters in the light of the heuristic information from the search procedure. Experiments and comparisons based on a suit of well-known test functions indicate that DSAHS achieves competitive results on the most of the test functions.