A genetic algorithm for designing optimal patch configurations in GIS

Abstract

Optimal patch design is a generic problem in which the objective is to find the best configuration of patches subject to multiple criteria. It is a hard problem in spatial geometry relevant to numerous applications in spatial planning and analysis. GIS packages do not have optimal patch design functions and there has been very little research in this area. This paper describes a computer system for solving optimal patch design problems in raster GIS. The method uses a genetic algorithm search heuristic combined with a region-growing programme that generates alternative patch configurations. The version described here solves multiple patch problems and was developed from an earlier version for designing single patches. In tests on a hypothetical planning problem the multi-patch version was found to be more efficient, but less effective, than the single patch version. The last part of the paper discusses outstanding issues regarding the applicability usability and external validity of the system and suggests ideas for further research.