Improving the social spider algorithm for constrained optimization

The optimization algorithms inspired from nature as intelligent optimization methods along with the classical methods have shown a considerable success and they have been used in solving many optimization problems in different areas such as engineering, industry, trade, and so on. In this paper we have considered the improvement of the genetic algorithm derived from a new social behavior of spiders to do an optimal search in global constrained optimization functions. The aim of improving this algorithm is accelerating the convergence while it can maintain the quality of good accuracy to achieve the optimal answer. the results of the proposed method on testing functions shows that the accuracy of the algorithm has fulfilled achieving the best answer in each test function along with the decline in the evaluation of optimization function (convergence ) faster than the method of social spider and it has achieved an average of more than 50 per cent improvement.

Keywords:

social spider algorithm , Constrained optimization , swarm intelligence , evolutionary algorithm , bio-inspired algorithm