A harmony search variant and a useful constraint handling method for the dynamic economic emission dispatch problems considering transmission loss

摘要

This paper proposes a new enhanced harmony search (NEHS) algorithm to solve dynamic economic emission dispatch (DEED) problems. NEHS is a variant of harmony search (HS) algorithm that is inspired by the music phenomenon. NEHS introduces two modifications to increase its competitiveness for the DEED problems. First, three levels of arbitrary distance bandwidths are used to enhance its global search and local search abilities. Second, both the best and the worst harmony vectors are selected to participate in the harmony memory considering operation in the second half of generations, which reaches a good compromise between improving solution quality and overcoming premature convergence. In addition, the improvised harmony vector at each generation is almost infeasible, and a repair technique is devised to drive it to the feasible domain efficiently. The performance of NEHS is evaluated on five DEED instances, and compared with that of the other four state-of-the-art HSs. Experimental results verify that NEHS is preferable in solving the five DEED instances as it can produce better solutions with less computation burden.