A Differential Evolution Algorithm with Adaptive Strategies for Constrained Optimization Problem

Constndned optimization problems are widely used in real-world applications as optimization models. Due to the complexity of the objective itself as well as too tight constraints, it is difficult to obtain the global optimal solution to these problems. In this manuscript, an improved differential evolutionary algorithm is proposed from the perspective of operator design and constraint handling. Firstly, in order to enhance the exploration ability of the algorithm, a heuristic mutation operator with better point information is constructed. Secondly, an improved dynamic epsilon constraint handling method is developed, in which the value of the epsilon decreases as the iteration number increases. The method can increase effectively the feasible individual in populations. Finally, the simulation results on 10 benchmark functions show that the proposed algorithm is effective and robust when compared with similar algorithms.