Session:

LBP - Late Breaking Papers

Title:

Constraint Handling of an Optical Components Selection Problem using a new Genetic Crossover Scheme

Authors:

Mohammad Amin
Malin Premaratne

Abstract:

This paper proposes a new crossover scheme called Controlled Content Crossover (CCC). CCC is applied to solve a commercially important optimization problem associated with the selection of the optimum set of components in optical network. CCC searches for the optimized solution and at the same time controls the value of the constrained parameter. This parameter is the optical component dispersion value in this paper. Therefore, by preserving the feasibility of the search points, CCC performs a genetic search over the feasible space. The simulation results show that the performance of the CCC is comparable with the answers obtained from CPLEX software and is better than the simple repair-based Genetic Algorithm proposed in the paper.