Future Generation Computer Systems
Artificial Intelligence in Geography
Artificial Intelligence in Geography
Geographical information systems and location science
Computers and Operations Research - Location analysis
An intelligent GIS-based spatial zoning system with multiobjective hybrid metaheuristic method
IEA/AIE'2004 Proceedings of the 17th international conference on Innovations in applied artificial intelligence
Ant colony system: a cooperative learning approach to the traveling salesman problem
IEEE Transactions on Evolutionary Computation
Ant system: optimization by a colony of cooperating agents
IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics
Ant intelligence for solving optimal path-covering problems with multi-objectives
International Journal of Geographical Information Science
| Hi-index | 0.00 |
This paper presents a new method to solve site selection problems using ant colony optimization (ACO) techniques. Optimal spatial search for siting public facilities is a common task for urban planning. The objective is to find N optimal sites (targets) for sitting a facility so that the total benefits are maximized or the total costs are minimized. It is straightforward to use the brute-force method for identifying the optimal solution by enumerating all possible combinations. However, the brute-force method has difficulty in solving complex spatial search problems because of a huge solution space. Ant colony optimization can provide a useful tool for site selection. In this study, the integration of ACO with geographic information systems is proposed to include various types of spatial variables in the optimization. A number of modifications have also been introduced so that ACO can fit spatial allocation problems. The novelty of this research includes the adoption of the strategies of neighborhood pheromone diffusion, tabu table adjusting, and multi-scale optimization. This method has been applied to the allocation of a hypothetical facility in Guangzhou City, China. The experiment indicates that the proposed model has better performance than the single search and the genetic algorithm for solving common site search problems.