Hybrid differential evolution harmony search algorithm for numerical optimization problems

To improve the optimization performance of harmony search algorithm, a hybrid differential evolution harmony search (HDEHS) algorithm is presented in this paper. In this algorithm, mutation and crossover operation are adopted instead of harmony memory consideration and pitch adjustment operation, which greatly improves the convergence rate. Moreover, the key parameters such as mutagenic factor and crossover rate are adjusted dynamically to balance the local and global search. Through several benchmark experiment simulations, the proposed algorithm has demonstrated stronger convergence and stability than the original harmony search algorithm and its typical improved algorithms reported in recent literatures.