Abstract: To solve the problem that the flexibility and solution space of cloud service composition were limited caused by determinate processing sequence in service composition,a matrix of process logic for describing the processing sequence was presented,and the service composition model for shortest time,lowest cost,minimal energy and highest reliability was established based on double-layer Ant Colony Optimization (ACO) algorithm.To avoid the conflicts between processing sequence and process logic of task,an interchange mechanism of processing sequence was put forward.The effectiveness of double-layer ACO was verified by the simulation,and the change rule of double-layer ACO's convergence rate was discussed under the different logistics condition and process logic.

Key words: double-layer ant colony algorithm, cloud manufacturing, service composition, process logic

摘要： 为解决在确定的子任务加工顺序条件下进行服务组合时,云服务组合的灵活性降低、解空间被限制等问题,提出了用于描述加工顺序的工艺逻辑矩阵,建立了基于双层蚁群算法的,以时间最短、成本最低、能耗最少、可靠性最高为目标的服务组合优选模型。针对加工顺序与任务工艺逻辑要求不一致的情况,在算法中加入了基于工艺逻辑的加工顺序交换机制。通过仿真实验验证了该算法的有效性,并研究了在不同物流条件和工艺逻辑要求下,算法收敛速度的变化规律。

关键词: 双层蚁群算法, 云制造, 服务组合, 工艺逻辑