Particle Swarm in Binary CSPs with Dynamic Variable Ordering

The variable ordering of constraint satisfaction problems affect the performance of search algorithms in CSPs. Dynamic Variable Ordering (DVO) has more advantage in improving the performance of search algorithms than static variable ordering. It is a newly developed method recent years that using particle swarm algorithm to solve binary constraint satisfaction problems, which is a global stochastic optimized algorithm making use of swarm to search the whole solution space, and each particle represents a candidate solution of the problem. The algorithm discovers a solution satisfying condition specified of the solution space by acting each other among these particles. We add the dynamic variable ordering to the particle swarm algorithm in constraint satisfaction problems by improving the evaluation function of the particle swarm algorithm, which enhances the searching efficiency of particle swarm algorithm in CSPs and finds the solution of CSPs faster. Kinds of random constraint satisfaction problem experiments indicated that our efforts were effective.