基于改进粒子群算法的组合服务优化

Optimization of Composition Service Based on Modified Particle Swarm Optimization

该文提出了一种基于遗传的粒子群算法来寻找最优组合服务。首先,设计了Web服务组合模型对组合服务进行形式化描述;接着,提出了候选服务筛选思想来解决组合数庞大问题;最后,引入遗传算法的交叉策略来避免粒子陷入停滞状态,同时采用了调整参数ω来改善粒子过早陷入局部最优。实验结果表明,基于改进粒子群优化的组合服务算法可靠、有效,能够获得综合Qos较好的解。

This paper proposes a Particle Swarm Algorithm based on agenetic to find the optimal combination of service. First,the design of Web service composition model to formally describe a combination of service.Second, for the huge number combinations of service problem, proposed the screening idea of candidate services. Finally, for the particle easily stalled issues, The paper introduced crossover strategy of Genetic Algorithm, while particle swarm easily fall into local optimum, used the adjusted parameter ω to improve. Experimental results show that the algorithm-Optimization of Composition Service Based on Modified Particle Swarm Optimization is reliable and effective, it is able to get a better solution which integrated Qos.