群体协作的果蝇优化算法及其在Web服务组合中的应用研究

发布时间：2018-01-08 02:15

本文关键词：群体协作的果蝇优化算法及其在Web服务组合中的应用研究　出处：《安徽大学》2016年硕士论文　论文类型：学位论文

【摘要】：果蝇优化算法(Fruit Fly Optimization Algorithm, FOA)是模仿果蝇在觅食过程中的合作行为而提出的一种新型群智能优化算法。FOA算法的原理是根据每只果蝇随机飞行到的位置计算各自的味道浓度值,并找出最佳味道浓度值,然后不断迭代,最终找到食物,实现优化问题的求解。FOA具有计算过程简单和易于理解的优点,然而FOA也存在一些缺点,例如,FOA使用固定步长容易导致算法的局部搜索能力和全局搜索能力失去平衡；FOA初始位置的选择对算法的稳定性造成了很大影响。本文针对FOA存在的缺点,对基本果蝇优化算法进行改进以提高算法的寻优性能,提出了群体协作的果蝇优化算法(Collaborative Swarm Fruit Fly Optimization Algorithm, CSFOA),并将CSFOA应用于求解Web服务组合问题中,本文的主要研究工作如下：(1)针对FOA的不足,本文提出一种群体协作的果蝇优化算法。首先,CSFOA采用双种群的协作机制和递减步长的策略,有效提高了算法的寻优精度和收敛速度。其次,CSFOA使用搜索系数h控制初始果蝇群体位置的选择以提高算法的收敛稳定性。(2)将CSFOA应用于连续型函数优化问题,并对18个经典的Benchmark函数进行测试,并与经典的群智能优化算法进行了大量对比。实验结果表明,CSFOA从整体上比FOA具有更好的全局搜索能力、更快的收敛速度、更高的收敛精度和稳定性。同时,与IFOA、PSO、DE相比,特别在高维函数求解方面,CSFOA具有更高的寻优精度和稳定性。(3)为更全面验证CSFOA的实际应用能力,将CSFOA应用于求解离散型的Web服务组合问题。通过果蝇的位置信息找到工作流中各个Web服务的位置,最后使用适应度函数计算组合服务质量的高低。将CSFOA的实验结果与FOA、PSO和DE的结果进行对比分析,实验结果证明,CSFOA具有更好的求解精度和求解速度；同时,稳定性上CSFOA优于PSO和DE。
[Abstract]:Fruit Fly Optimization Algorithm. FOAA). A new swarm intelligence optimization algorithm, FOA algorithm, is proposed to simulate the cooperative behavior of Drosophila during foraging. The principle of FOA algorithm is to calculate the taste concentration of each fly based on the random flight position. And find out the best taste concentration, and then iterate, finally find food, to achieve the optimization problem solving. FOA has the advantages of simple and easy to understand the calculation process, but FOA also has some shortcomings. For example, using fixed step size of FOA can lead to the imbalance of local search ability and global search ability of the algorithm. The selection of the initial position of FOA has a great influence on the stability of the algorithm. Aiming at the shortcomings of FOA, the basic Drosophila optimization algorithm is improved to improve the performance of the algorithm. A collaborative Swarm Fruit Fly Optimization Algorithm is proposed. CSFOAA, and applies CSFOA to solve the Web service composition problem, the main research work of this paper is as follows: 1) aiming at the shortage of FOA. In this paper, we propose an optimization algorithm for Drosophila. Firstly, CSFOA adopts the cooperation mechanism of two populations and the strategy of decreasing step size, which can effectively improve the optimization accuracy and convergence speed of the algorithm. CSFOA uses the search coefficient h to control the initial Drosophila population location to improve the convergence stability of the algorithm. The CSFOA is applied to the continuous function optimization problem. The 18 classical Benchmark functions are tested and compared with the classical swarm intelligence optimization algorithm. The experimental results show that. CSFOA has better global search ability, faster convergence speed, higher convergence accuracy and stability than FOA. At the same time, compared with FOA PSODE. Especially in the aspect of solving high-dimensional function, CSFOA has higher precision and stability of optimization. It is more comprehensive to verify the practical application ability of CSFOA. The CSFOA is applied to solve the discrete Web service composition problem, and the location of each Web service in workflow is found by the location information of Drosophila. Finally, the fitness function is used to calculate the quality of service composition. The experimental results of CSFOA are compared with those of CSFOA and DE, and the experimental results are proved. CSFOA has better accuracy and speed. At the same time, the stability of CSFOA is better than that of PSO and De.
【学位授予单位】：安徽大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP18;TP393.09

【相似文献】