信息管理软件自动化测试方法的设计  

Design of Automated Testing Method for Information Management Software

作　　者：裘雨音 张静 陈江尧 QIU Yuyin;ZHANG Jing;CHEN Jiangyao(State Grid Zhejiang Electric Power Co.,Ltd.,Hangzhou 310012,China;Training Center,State Grid Zhejiang Electric Power Co.,Ltd.,Hangzhou 310015,China;Zhejiang Huayun Information Technology Co.,Ltd.,Hangzhou 310012,China)

机构地区：[1]国网浙江省电力有限公司,浙江杭州310012 [2]国网浙江省电力有限公司培训中心,浙江杭州310015 [3]浙江华云信息科技有限公司,浙江杭州310012

出　　处：《自动化仪表》2024年第1期86-90,共5页Process Automation Instrumentation

摘　　要：为了提高信息管理软件异常识别的准确率,基于深度神经网络设计了一种信息管理软件自动化测试方法。利用Markov链构建信息管理软件任务剖面模型,提取用于信息管理软件自动化测试的任务流。深度优先遍历剖面,融合自动化测试操作行为的变化条件、后置条件、操作概率等信息。在此基础上,提取信息管理软件自动测试的任务流。结合已知的软件任务状态转移概率计算极限概率,将重要性显著的用例与参数信息作为优先测试的信息,完成网络训练。建立软件异常自动化识别模型,根据异常识别结果对软件异常位置进行定位,完成信息管理软件自动化测试。测试结果表明,该方法的软件自动化测试结果更准确、自动化效率更高,具有一定的应用价值。To improve the accuracy of anomaly identification of information management software,an automated testing method for information management software is designed based on deep neural network.Markov chain is utilized to construct an information management software task profile model to extract the task flow for automated testing of information management software.Deep priority traverses the profile to fuse the information of change conditions,post conditions,and operation probabilities of automated test operation behaviors.On this basis,extract the task flow for automated testing of information management software.Combine the known software task state transfer probability to calculate the limit probability,and take the use cases and parameter information with significant importance as the information of priority testing to complete the network training.Establish the software anomaly automation identification model,locate the software anomaly location according to the anomaly identification results,and complete the information management software automation test.The test results show that the software automation test results of this method are more accurate,and the automation efficiency is higher,which has certain application value.

关 键 词：深度神经网络 信息管理 软件 自动化 测试 自编码器 

分 类 号：TH-3[机械工程]