Improving the Music Harmony Search Using a Differential Exploration Phase for Solving Bound Constrained Optimization Problems

Very frequently, when solving a complex optimization problem, one is confronted that it comes from an NP-hard class, which means high difficulty to get the optimal solution. As an approximate optimization technology, meta-heuristics are good substitutions for conventional mathematics-based deterministic technology because it can provide acceptable solutions in reasonable computational time. As one of the classical meta-heuristics technology, harmony search (HS), inspired by music composition process, shows effectiveness in solving NP-hard complex optimization problems in a variety of research and practical fields, while it has a unique iteration paradigm. However, in recent 2 decades of years, the theoretical test and researches on harmony search are not sufficiently conducted. In this paper, we propose an improved harmony search (iHS) with a differential exploration phase and test its effectiveness and efficiency on the 2024 congress of evolutionary computation (CEC2024) competition benchmark test suite. Comparing with various classical and new proposed meta-heuristics, The experimental results demonstrate that iHS possesses better performance which shows its considerable potentials to be adequately developed in the future.