Hybrid Constrained Evolutionary Algorithm for Numerical Optimization Problems

Constrained optimization are naturally arises in many real-life applications, and is therefore gaining a constantly growing attention of the researchers. Evolutionary algorithms are not directly applied on constrained optimization problems. However, different constraint-handling techniques are incorporated in their framework to adopt it for dealing with constrained environments. This paper suggests an hybrid constrained evolutionary algorithm (HCEA) that employs two penalty functions simultaneously. The suggested HCEA has two versions namely HCEA-static and HCEA-adaptive. The performance of the HCEA-static and HCEA-adaptive algorithms are examined upon the constrained benchmark functions that are recently designed for the special session of the 2006 IEEE Conference of Evolutionary Computation (IEEE-CEC'06). The experimental results of the suggested algorithms are much promising as compared to one of the recent constrained version of the JADE. The converging behaviour of the both suggested algorithms on each benchmark function is encouraging and promising in most cases.