HYBRID METHOD BASED ON EXPONENTIAL PENALTY FUNCTION AND MOMA-PLUS METHOD FOR MULTIOBJECTIVE OPTIMIZATION

Abstract

In this paper, we propose a modified version of the MOMA-plus method to solve multiobjective optimization problems. We use an exponential penalty function instead of the Lagrangian penalty function in the initial version of MOMA-Plus in order to improve the convergence and distribution to the Pareto optimal solutions. The theoretical and numerical results show that this new version improves the quality of the obtained solutions compared to the last version. Six test problems have been successfully resolved, allowing us to highlight the good convergence and good distribution of Pareto optimal solutions.