摘要
针对软件可靠性模型中参数估计不精确的问题,提出了一种基于和声搜索算法的软件可靠性模型参数估计方法.为了避免和声搜索算法在求解参数时陷入局部最优解,对算法作了改进:在和声记忆库初始化时采用反向学习策略,提高了收敛速度;利用全局信息产生新解,提高了全局搜索能力.使用该方法对5组数据的两个软件可靠性模型的参数进行了估计,实验结果表明,本算法应用于参数估计具有可行性和有效性,在精度和算法的收敛性上,明显优于其他智能算法.
As to the problem of inaccurate estimation of the parameters of software reliability model, a method of estimating parameters of software reliability model based on the harmony search algorithm is proposed in this paper. In order to avoid the harmony search algorithm for solving the parameters into a local optimal solution, two improvements are made. Firstly, the op- position-based learning is used to initialized harmony memory to improve the speed of convergence. Secondly, global information is used to generate the new to improve the global search ability. The parameters of two software reliability models of seven data sets were estimated by the proposed method. The experimental results show that the algorithm is feasibility and effectiveness in parameter estimation and is superior to other meta-heuristic algorithms in the convergence and accuracy.
出处
《山东理工大学学报(自然科学版)》
CAS
2017年第2期44-48,共5页
Journal of Shandong University of Technology:Natural Science Edition
基金
安徽省教育厅自然科学基金重点项目(KJ2014A100
KJ2016A308)
关键词
软件可靠性模型
和声搜索算法
参数估计
software reliability model
harmony search algorithm
parameter estimation