An Infeasibility Objective for Use in Constrained Pareto Optimization
- 3.7k Downloads
A new method of constraint handling for multi-objective Pareto optimization is proposed. The method is compared to an approach in which each constraint function is treated as a separate objective in a Pareto optimization. The new method reduces the dimensionality of the optimization problem by representing the constraint violations by a single “infeasibility objective”. The performance of the method is examined using two constrained multi-objective test problems. It is shown that the method results in solutions that are equivalent to the constrained Pareto optimal solutions for the true objective functions. It is also concluded that the reduction in dimensionality of the problem results in a more transparent set of solutions. The method retains elegance of the underlying Pareto optimization and does not preclude the representation of a constraint as an objective function where this is considered important. The method is easily implemented and has no parameters to be tuned.
KeywordsMultiobjective Optimization Pareto Optimal Solution Constraint Violation Pareto Optimization Pareto Solution
Unable to display preview. Download preview PDF.
- Fonesca C.M., Fleming P.J.: Multiobjective Genetic Algorithms Made Easy: Selection, Sharing and Mating Restriction. Genetic Algorithms in Engineering Systems: Innovations and Applications Conference Publication No.414, IEE, (1995), 12–14.Google Scholar
- Goldberg D.E., Richardson J.: Genetic Algorithms with sharing for multi-modal function optimization. Proceedings of the 7th International Conference on Genetic Algorithms’, (1987), 41–49.Google Scholar
- Deb K., Goldberg D.E.: An investigation of Niche and Species Formation in Genetic Function Optimization. Proceedings Of The Third International Conference On Genetic Algorithms, (1989), 42–50.Google Scholar