应用于服务组合的改良粒子群算法

Improved Particle Swarm Optimization for Service Composition

摘要: 当今社会，互联网、物联网、云计算以及大数据的快速发展和普及使得网络应用越来越广，越来越多的行业和相应的网络进行越来越深地融合。许多原本属于线下的服务经过封装后被搬上了网络上，同类型的服务也越来越多。如何在互联网大数据环境下快速找到满足用户个性化需求的服务组合已经成为亟需解决的问题。为了解决这个问题，本论文提出了一种改良粒子群服务组合方法，此改良粒子群算法根据服务组合问题的特点分别从四个方面加入了逃出局部最优的机制，根据用户的服务组合请求去快速组合出更优的服务组合方案。本文通过实验与其他相关的服务组合优化算法在最优性、时间复杂度以及收敛性三个指标进行了对比。根据实验结果分析，本文所提出的方法在最优性、时间复杂度以及收敛性三个方面整体上表现出来良好的性能。

Abstract: In today’s society, the rapid development and popularization of the Internet, Internet of Things, cloud computing, and big data make network applications increasingly widely used. And increasing industries and corresponding networks are integrated increasing deeply. Many services that originally belonged to the line have been packaged and put on the Internet, and there are increasing services of the same type. How to quickly find a service combination that meets the personalized needs of users in the Internet big data environment has become an urgent problem to be solved. In order to solve this problem, this paper proposes an improved particle swarm service composition method. According to the characteristics of the service composition problem, the improved particle swarm algorithm adds a mechanism of escaping from the local optimum from four aspects. According to a user’s service composition request, quickly compose a better service composition solution. This paper compares the optimality, time complexity and convergence with other related service composition optimization algorithms through experiments. According to the analysis of the experimental results, the method proposed in this paper shows good performance on the whole in terms of optimality, time complexity and convergence.