基于复杂网络理论的高铁列控系统危险成因动力学建模及关键致因辨识

System hazard analysis and key causal factors identification are basis of the safety guarantee of High-speed Railway Train Control (HRTC) systems. Existing researches tend to utilize the systemic methods. However, due to without considering temporal relations between causal factors and the limitation of non-linear interactions, it is difficult to capture the key factors effectively. Firstly, this project will begin with designing a new control action temporal logic for HRTC systems, based on which logic operation will be utilized to analyze the hazard causal factors and their temporal relations. Secondly, the complex network theory will be utilized to model causation networks and the dynamics of the evolutionary processes of hazards in the causal factors networks, where the mechanisms about how causal factors lead to hazards could be revealed. Finally, with the topological characteristics of causation networks, the dynamic model is utilized to analyze the sensitivity and influence of the causal factors to capture the key causal factors. The research results from this project will enrich the existing system hazard causation analysis approaches, and provide theoretical foundations for the upgrade and optimization of the actual safety guarantee deployments of HRTC systems.

系统危险分析及关键致因辨识是高铁列控系统安全保障研究的基础。现有研究趋于采用“系统观”的分析方式，但由于未考虑致因时序关系并且受致因非线性作用的限制，难以有效实现对关键致因的辨识。本项目将首先从设计面向高铁列控系统的控制时序逻辑出发，利用数理逻辑运算实现对系统危险致因及其时序关系的分析。其次，借助复杂网络理论构建致因关系网络，对危险成因动力学过程进行建模，从而揭示致因触发系统危险的动力学作用机理。最后，结合致因网络的拓扑特征，利用危险成因动力学模型分析致因导致危险时的敏感程度和影响程度，从而实现对关键致因的辨识。本项目的研究不仅能作为现有系统危险致因分析方法的必要补充，也将为高铁列控系统安全保障部署的升级和优化提供新的理论依据。

项目针对现有高铁列控系统危险致因分析工作难以对关键致因辨识的问题展开相关研究，提出了面向高铁列控系统扩展的统一建模语言，实现系统分层控制结构的半形式化建模；分析了高铁列控系统控制行为时序特征，提出了面向高铁列控系统的控制时序逻辑，实现对危险致因时序关系的分析；利用复杂网络理论构建了危险致因关系网络，提出了基于网络级联失效过程的系统危险成因动力学模型，揭示了致因触发系统危险的动力学作用机理；分析了致因网络的实际拓扑特征，结合成因动力学模型提出了高铁列控系统危险关键致因辨识方法；利用CTCS-3级高铁列控系统典型运营场景验证了本项目模型与方法的效果。本项目研究成果不仅为现有系统危险致因分析方法提供必要补充，也为高铁列控系统安全保障部署的升级和优化提供了新的理论依据。