An Improved Optimization Approach Based on Competitive Algorithm

In the paper, we developed an improved approach for the imperialist competitive algorithm to help the original algorithm escape from local optimal area and obtain better performance. The Nelder-Meand simplex method is also introduced to execute alternately with the original procedures of the algorithm. The proposed approach is tested on twelve widely-used benchmark functions to check its validity. Compared with other relative studies, the proposed approach shows faster convergence rate, better search ability, and better stability than the original algorithm and other relative methods.