解空间搜索是约束求解的关键环节,目前较为常用的搜索算法一般是基于二元约束或单一搜索策略设计的,本文设计了六个基于多元约束的混合搜索算法(BM_GASBJ,BM_GBJ,BM_CBJ,FC_GASBJ,FC_GBJ,FC_CBJ),它们分别混合同一类搜索策略中不同算法或不同类搜索策略;分析并给出了不同混合算法的性能差异,系统测试结果表明混合搜索算法明显提高了解搜索效率和约束求解系统的性能。
Searching is the key step of constraint solving. Most of search algorithms are based on binary constraint or designed upon a single search strategy. Six hybrid search algorithms (BM_GASBJ, BM_GBJ, BM-CBJ, FC_GASBJ, FC_GBJ, FC_CBJ) based on non-binary constraint are illustrated. The hybrid algorithms are blended by various search algorithms or strategies. The differences of these hybrid algorithms are pointed out and analyzed . The constraint solving system we designed implements all of the hybrid algorithms. The results indicate that the new algorithms have a higher searching efficiency and improve the performance of the constraint solving system.