Surrogate-Based Optimization for Complex Engineering problems

This paper proposes a surrogate modeling-based optimization approach for solving complex engineering optimization problems. The main challenge is to use the surrogate model for evaluating computationally expensive and constrained problems. Kriging and radial basis function models are considered for modeling both performances and constrains. Penalty technique is considered for dealing with constraints. To show the efficiency of the proposed algorithms, obtained results are compared with the conventional equation-based particle swarm optimization (PSO) algorithm results. Accuracy and robustness of the approaches are also demonstrated. Experimental results indicate that the proposed approaches are promising for solving complex constrained optimization problems.