Solving permutation constraint satisfaction problems with Artificial Ants

摘要: We describe in this paper Ant-P-solver, a generic constraint solver based on the Ant Colony Optimization (ACO) meta-heuristic. The ACO metaheuristic takes inspiration observation of real ants collective foraging behaviour. idea is to model problem as search best path graph. Artificial walk trough graph, stochastic and incomplete way, searching for good paths. communicate local indirect by laying pheromone trail edges graph. Ant-P-solver has been designed solve general class combinatorial problems, i.e., permutation satisfaction goal which find n known values, be assigned variables, under some constraints. Many problems involve such global Ant-P-solver capabilities are illustrated, compared with other approaches, three these n-queens, all-interval series car sequencing problems.