QoS全局最优的多目标Web服务选择算法  

Global QoS optimizing and multi-objective Web service selection algorithm

作　　者：方其庆[1,2] 刘庆华[2] 彭晓明[3] 胡亚慧[2] 

机构地区：[1]通信指挥学院,武汉430010 [2]空军雷达学院,武汉430019 [3]武汉大学计算机学院,武汉430072

出　　处：《计算机应用研究》2009年第12期4442-4445,4448,共5页Application Research of Computers

基　　金：国家"863"计划资助项目(2007AA01Z138)

摘　　要：针对现有方法的不足,提出一种基于QoS全局最优的多目标动态W eb服务选择算法。在给出动态服务组合模型的基础上,以"抽象服务规划"为输入,以用户的非功能性需求为全局约束,将动态服务选择问题转换为一个带QoS约束的多目标服务组合优化问题;利用多目标蚁群算法,多个目标函数被同时优化并产生一组满足约束条件的Pareto优化解。通过运用实验与基于多目标遗传算法的W eb服务选择算法进行对比,证明了该方法的可行性和有效性。To resolve the deficiency of the existing approaches, this paper presented a novel global QoS optimizing dynamic Web service selection algorithm. Firstly, proposed a model of Web service composition. Consequently,with the ＂abstract service plan＂ as input and the user non-functlonal requirements as global constraints, converted the Web service selection problem into a multi-objective optimization problem. Furthermore, based on the multi-objective ant colony algorithm（ MOACO）, the proposed algorithm simultaneously optimized the different QoS parameters of the workflow. In this way, calculated a set of optimal solutions, known as Pareto set. Experimental results prove the proposed algorithm outperforms the recently published QoS global optimization based on multi-objective genetic algorithm （MOGA） , specially designed for solving the Web services selection problem.

关 键 词：WEB服务 服务选择 全局优化 多目标蚁群算法 

分 类 号：TP393[自动化与计算机技术—计算机应用技术] TP301.6[自动化与计算机技术—计算机科学与技术]