摘要
软件可靠性模型参数估计的结果决定了软件可靠性预测的准确度。由于现在的大多数软件可靠性模型都是非线性的,因此其参数估计是比较困难的。在此基础上,论文提出基于狼群算法的软件可靠性模型参数估计方法,结合极大似然估计方法来构造适应值函数,在算法执行过程中剔除错误的解并加入先验知识提高解的精确度,利用先验知识来优化参数的搜索方向。文章使用来自实际工业界的5组数据集来估计GO模型的参数,并进行预测和比较。仿真实验结果验证了此方法进行参数估计的准确度更高,优化性能更好,具有更好的模型预测效果。
The results of the software reliability model parameter estimation determine the accuracy of the software reliability prediction.Since most software reliability models are nonlinear,their parameter estimation is difficult.On this basis,this paper pro⁃poses a software reliability model parameter estimation method based on wolf group algorithm,combined with the maximum likeli⁃hood estimation method to construct the fitness value function,eliminates the wrong solution in the algorithm execution process and adds the prior knowledge to improve the solution.Prior knowledge is used to optimize the search direction of the parameters.The arti⁃cle uses five sets of data sets from the real industry to estimate the parameters of the GO model and make predictions and compari⁃sons.The simulation results verify that the proposed method has higher accuracy,better optimization performance and better model prediction.
作者
于苗苗
朱兵
李震
王东升
魏海峰
YU Miaomiao;ZHU Bing;LI Zhen;WANG Dongsheng;WEI Haifeng(School of Electronics and Information,Jiangsu University of Science and Technology,Zhenjiang 212003;Shanghai Merchant Ship Design and Research Institute,Shanghai 201203;School of Computer Science,Jiangsu University of Science and Technology,Zhenjiang 212003)
出处
《计算机与数字工程》
2020年第12期2948-2953,共6页
Computer & Digital Engineering
关键词
狼群算法
软件可靠性模型
参数估计
模型预测
wolf pack algorithm
software reliability model
parameter estimation
model prediction
