Multi-HM Adaptive Harmony Search Algorithm and its Application to Continuous Function Optimization

An improved harmony search algorithm called Multi-HM Adaptive Harmony Search (MAHS) algorithm is investigated. The new harmonies of multiple harmony memories compete each other to produce a better one assigned as the new harmony for both HMs. Adaptive PAR is employed and the new harmony is introduced into pitch adjusting bandwidth. The golden section is adopted as the border of pitch adjusting. The experimental results show MAHS performs well for both low-dimensional and high-dimensional problems when applied to five common benchmark functions and compared with HS, IHS and GHS.