Sizing optimization of skeletal structures with a multi-adaptive harmony search algorithm

The harmony search (hs) algorithm is a popular metaheuristic optimizationmethod that reproduces the music improvisation process in searching for a perfect state ofharmony. hs has a remarkable ability in detecting near global optima at low computationalcost but may be ineective in performing local search. this study presents the multi-adaptive harmony search (mahs) algorithm for sizing optimization of skeletal structureswith continuous or discrete design variables. the main dierence between the proposedalgorithm and classic hs is the way of choosing and adjusting the bandwidth distance(bw). furthermore, mahs dynamically updates the harmony memory considerationrate (hmcr) and pitch-adjusting rate (par) parameters during the search process.the robustness and performance of the mahs algorithm are evaluated in comparison withliterature, and in particular, with well-known hs variants such as global-best harmonysearch (ghs), self-adaptive harmony search (sahs), and ecient harmony search(ehs). optimization results obtained by the mahs algorithm conrm the validity of theproposed approach.