结合蜂群算法与RBF神经网络的软件错误定位方法  

Software Fault Localization Based on Artificial Bee Colony Algorithm and RBF Neural Network

作　　者：吕琼帅[1] 史玉珍[1] 夏栋梁[1] Lv Qiongshuai;Shi Yuzhen;Xia Dongliang(Software College,Pingdingshan University,Pingdingshan 467000,China)

机构地区：[1]平顶山学院软件学院,河南平顶山467000

出　　处：《科技通报》2017年第4期164-167,共4页Bulletin of Science and Technology

基　　金：河南省科技攻关计划项目(KJT142102210226);河南省高等学校青年骨干教师资助计划资助项目(183);平顶山学院青年基金项目(PXYQNJJ2016013)

摘　　要：软件错误是软件开发过程中经常出现的问题,有效地定位错误能够提高软件开发的效率。通过对软件运行时信息的收集与分析,从机器学习的角度提出了一种将蜂群算法与RBF神经网络结合的软件错误定位模型。首先,利用蜂群算法克服RBF神经网络由于隐层节点中心值随机化而导致的网络性能较差的问题,对RBF神经网络模型进行优化与改进。然后,对程序运行过程中的相关信息进行编码,将此编码作为蜂群神经网络模型的训练数据集。最后,根据虚拟测试集进行软件错误定位,实验过程中选用Siemens Suite套件进行了仿真实验。实验结果表明,结合蜂群算法优化与RBF神经网络的软件错误定位方法在软件错误定位方面能够表现出较好的性能。Software error is a frequent problem in software development process,which can improve theefficiency of software development.By collecting and analyzing the information of the running time of thesoftware,a software error location model is proposed,which is based on the machine learning.First,theRBF neural network model is optimized and improved by using the swarm algorithm to overcome theproblem that the network performance is poor because of the hidden layer node center valuerandomization.Then,encoding will be used as the training data set of the bee colony neural networkmodel.Finally,according to the virtual test set,the software error is located,and the simulationexperiment is carried out with Suite Siemens.The experimental results show that the RBF neural networkoptimized by swarm optimization algorithm has good performance in software error localization.

关 键 词：错误定位 蜂群算法 RBF神经网络 可疑度 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论] TP183[自动化与计算机技术—计算机科学与技术]