基于BPSO算法的Web系统参数优化研究

发布时间：2018-06-08 09:54

本文选题：Web系统 + 参数优化　；参考：《华南理工大学》2014年硕士论文

【摘要】：互联网深刻影响着我们的生活，日益增长的用户流量给我们的Web系统提出了挑战，要求我们充分利用Web系统的性能服务更多的用户。通过调整Web系统的参数配置可以显著地提升系统性能，但是由于Web系统的参数众多，人工调整参数配置相当麻烦，而且需要工作人员有丰富的经验。本文通过分析自建Web系统，综合各方面条件，利用二进制粒子群（BinaryParticle Swarm Optimization，BPSO）算法及其改进算法优化Web系统的参数配置，寻找最优参数配置，使Web系统性能表现最优。本文主要包括以下内容：（1）实验平台部署。经过分析，选择研究由Apache、MySQL和PHP组成的Web系统，选择ApacheBench作为性能测试工具。采用Python编写控制系统，实现Web系统和性能测试工具的连接，，完成实验平台部署。（2）BPSO算法实现Web系统参数优化。深入分析BPSO算法的基本原理和工作流程以及影响算法性能的因素，把它实际应用到Web系统参数优化问题中。针对Web系统的主要可调配置参数进行二进制编码，采用计算机随机的方法获得初始种群，通过BPSO算法迭代优化，获得全局最优解，给出实验结果和算法性能分析。（3）改进BPSO算法，并用其实现Web系统参数优化。通过分析，BPSO算法容易早熟，前期粒子的多样性过快降低，后期局部搜索能力弱。本文在算法早期通过引入耗散操作来增加粒子的多样性，在算法后期通过引入爬山算法来增强算法的局部搜索能力，形成新的混合算法。用新的混合算法进行Web系统参数优化，获得全局最优解，给出实验结果和算法性能对比分析。在有限的资源和负载下，本文通过部署实验平台进行Web系统性能测试实验，运行算法程序找到系统最优或接近最优的参数配置，给出了实验结果和分析，验证优化算法的有效性，有实际的应用价值。
[Abstract]:The Internet has a profound impact on our daily life. The increasing user traffic challenges our Web system and requires us to make full use of the performance of the Web system to serve more users. By adjusting the parameter configuration of the Web system, the performance of the system can be significantly improved, but because of the large number of parameters of the Web system, the manual adjustment of the parameters is quite troublesome, and the staff is required to have rich experience. In this paper, the binary Particle Swarm Optimization (BPSO) algorithm and its improved algorithm are used to optimize the parameter configuration of the Web system and to find the optimal parameter configuration to optimize the performance of the Web system. This paper mainly includes the following contents: 1) deployment of experimental platform. After analysis, the Web system composed of Apache MySQL and PHP is studied, and Apache Bench is chosen as the performance testing tool. The control system is programmed by Python to realize the connection between the Web system and the performance testing tools, and the deployment of the experimental platform is completed, and the BPSO algorithm is used to optimize the parameters of the Web system. The basic principle and workflow of BPSO algorithm and the factors that affect the performance of BPSO algorithm are analyzed in detail, and the BPSO algorithm is applied to the optimization of Web system parameters. The binary coding for the main adjustable configuration parameters of Web system is carried out. The initial population is obtained by computer random method. The global optimal solution is obtained by iterative optimization of BPSO algorithm. The experimental results and performance analysis of the algorithm are given. The improved BPSO algorithm is improved. It is used to optimize the parameters of Web system. By analyzing that the BPSO algorithm is easy to prematurely reduce the diversity of particles too quickly and the local search ability is weak in the later stage. In this paper, dissipative operations are introduced to increase the diversity of particles in the early stage of the algorithm, and a new hybrid algorithm is formed by introducing the mountain climbing algorithm to enhance the local search ability of the algorithm. The new hybrid algorithm is used to optimize the parameters of the Web system, and the global optimal solution is obtained. The experimental results are compared with the performance of the algorithm. Under the limited resources and load, the performance of the Web system is tested by deploying the experimental platform. The algorithm program is run to find the optimal or near optimal parameter configuration of the system. The experimental results and analysis are given to verify the effectiveness of the optimization algorithm and it has practical application value.
【学位授予单位】：华南理工大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP393.09;TP18

【参考文献】