摘要
随着信息物理融合系统(CPS)日益复杂,很难捕获其相关行为并以一种可解释的方式检测它.因此本文提出了一种基于线性支持向量机(LSVM)框架的规范挖掘算法,从有限长度的信号轨迹中挖掘具有可解释性的信号时序逻辑(STL)公式.挖掘出的STL公式可以作为监控行为的抽象,用于CPS运行时验证的监控机制中.该算法根据一组被标记的有限时间轨迹和STL模板公式,将信号时序逻辑特有的鲁棒性满意度与LSVM的优化算法相结合生成所需规范.另外,本文还将对一维信号轨迹的规范挖掘技术扩展到多维信号上.最后,通过两个案例研究来说明提出算法的可行性和优势,结果表明该算法在保证高准确率的前提下,执行效率和可解释性都优于其他的规范挖掘算法.
ion of monitoring behavior in the monitoring mechanism of CPS runtime verification.The algorithm combines the robustness satisfaction unique to signal temporal logic with the optimization algorithm of LSVM to generate the desired specification based on a set of labeled finite-time trajectories and the STL template formula.In addition,this paper extends the specification mining technique for one-dimensional signal traces to n-dimensional signal traces.Finally,two case studies are presented to illustrate the viability and advantages of the proposed algorithm,which is shown to be superior to other specification mining algorithms in terms of execution efficiency and interpretability while ensuring a high accuracy rate.
作者
刘峰
曹子宁
王福俊
李振
LIU Feng;CAO Zining;WANG Fujun;LI Zhen(College of Computer Science and Technology,Nanjing University of Aeronautics and Astronautics,Nanjing 211106,China;Key Laboratory of Safety-Critical Software(Nanjing University of Aeronautics and Astronautics),Ministry of Industry and Information Technology,Nanjing 211106,China;Science and Technology on Electro-optic Control Laboratory,Luoyang 471023,China)
出处
《小型微型计算机系统》
CSCD
北大核心
2024年第1期9-15,共7页
Journal of Chinese Computer Systems
基金
航空科学基金项目(20185152035)资助.
关键词
机器学习
规范挖掘
运行时验证
信息物理融合系统
信号时序逻辑
machine learning
specification mining
runtime verification
cyber-physical systems
signal temporal logic
