基于协同效应的并行制造云服务组合算法

Parallel manufacturing cloud service composition algorithm based on collaborative effect

为了解决并行结构下制造云服务的组合优化问题,从协同效应的角度提出一种基于反向和局部学习的蚁群算法。该算法以服务质量作为信息素构建协同效应评价模型,求解并行服务的协同效应值,作为启发函数参数。利用协同关系矩阵计算组合协同效应值,并将其与组合服务质量相结合,求解综合评价值最高的服务组合。实验结果表明,该算法能有效求解并行制造云服务的组合优化问题,并能较快地得到最优解。

To solve the problem of manufacturing cloud service composition optimization in parallel structure,an improved ant colony algorithm based on reverse-learning and local-learning was proposed from the perspective of collaborative effect.An evaluation model was established to calculate the value of collaborative effect as the heuristic function parameter,and the quality of service as the pheromone.Then,the collaborative effect of service composition calculated by collaborative relationship matrix was combined with the quality of service composition to obtain the optimal service composition.The experimental results showed that the improved algorithm was effective and feasible,which could achieve the global optimal solution quickly.