A New Evolutionary Algorithm for Constrained Optimization Problems

Abstract

In constrained optimization problems, evolutionary algorithms often utilize a penalty function to deal with constraints, which is, however, difficult to control the penalty parameters. To overcome this shortcoming, this paper presents a new constraint handling scheme. Firstly, a new fitness function defined by this penalty function and the objective function is designed. The new fitness function not only can classify all individuals in current population into different layers automatically, but also can distinguish solutions effectively from different layers. Meanwhile, a new crossover operator is also proposed which can produce more high quality individuals. Based on these, a new evolutionary algorithm for constrained optimization problems is proposed. The simulations are made on five widely used benchmark problems, and the results indicate the proposed algorithm is effective.