Constraint Handling with Modified Hypervolume Indicator for Multi-objective Optimization Problems

摘要

Many problems across various domains of research may be formulated as a multi-objective optimization problem. The Multi-objective Evolutionary Algorithm framework (MOEA) has been applied successfully to unconstrained multi-objective optimization problems. This work adapts the modified Hypervolume Indicator to incorporate constraints when used within the MOEA framework. Empirical results from a sample problem showed that the algorithm is capable of generating a high percentage of feasible solutions, while the shape parameter used to govern the desirability function make a trade-off between feasibility and Hypervolume. Furthermore, the shape parameter is shown to heavily influence the feasible solution's final goodness.