A Fast Global Group Search Optimizer algorithm
Kang Zhang, Xingsheng Gu · 2014
The Group Search Optimizer(GSO) is a novel optimization algorithm, which is inspired by searching behavior of animals. In this paper, we proposed an improved GSO algorithm named Fast Global Group Search Optimizer(FGGSO) to increase searching speed and balance the exploitation and exploration of the algorithm, which is based on our previous works. At first time, considering the complexity and time-consuming design of the producer's angle searching strategy, a novel local search mechanism, named campaign strategy, is developed, which is inspired by competition and cooperation between candidates in an electoral process. After that, a reconstruction operation is applied in searching process to guarantee the avoidance of the local minimum. The algorithm is evaluated on a set of 11 numerical optimization problems and compared favorably with other version of GSOs. Experimental results indicate the remarkable improvement on the performance of these problems.