关键词: 复杂任务, 竞争关系, 协同效应, 双边匹配, 改进粒子群算法

Abstract: To improve the stability of matching scheme and reduce the complexity of cloud task matching,according to the construction of task competition network and service collaboration network,a one to many two-sided matching problem based on competition and synergy effect was put forward.The expected utility theory was applied to calculate the satisfaction of both parties,and the aggregation methods of competition relationship based satisfaction between tasks and social network based satisfaction between services were proposed respectively.By taking the maximum of task satisfaction,service satisfaction,satisfaction between tasks and satisfaction between services as the goal,a two-way matching multi-objective optimization model was built,and the best solution was acquired by using the improved non dominated particle swarm algorithm and weighted TOPSIS.The automobile manufacturing experiments validated the effectiveness of the proposed model and algorithm.Compared with the three two-way models that were traditional,only considered task competition realtion and considered service synergy effect,the solution difference of four matching models was analyzed in various application scenarios.It was proved that the proposed model had the advantage in competitive and collaborative environment,and its optimal satisfactory stable matching was conformed to the real situation.

Key words: complex task, competition relationship, synergy effect, two-sided matching, improved particle swarm algorithm