A Hybrid Particle Swarm Optimization with Crossover and Mutation of Genetic Algorithm for Solving the Wide Constraint Problem

When optimizing the spatial data, a lot of constraints should be handled. Some constraints might be too wide for a metaheuristic algorithm, e.g. particle swarm optimization, to allocate the candidate locations outside a wide constraint. However, particle swarm optimization notably has fast computation characteristic and many researchers used this method for optimizing their spatial data. In the other hand, genetic algorithm has not only better exploitation-characteristic performance in searching but also has mutation and crossover that was proven in this study can be overcome the wide constraint problem. To minimize the drawback of genetic algorithm, i.e. need many computation resources, the hybrid particle swarm optimization with genetic algorithm through the use of crossover and mutation was used. Half of lower fitness values from particle swarm optimization were optimized using crossover and mutation in genetic algorithm. After merging the results of both methods, the optimum location showed that the proposed method was able to allocate the land use in a case study area outside the wide constraint.