考虑复杂关联情境的城市典型生命线运行风险因素识别方法研究

城市典型生命线以供热、电力、燃气系统为代表，具有公共性强、风险性高、关联性显著的特点。研究城市典型生命线运行风险因素识别问题对于相关部门防范风险具有重要意义，但现实中由多系统、多风险因素以及各类关联形成的复杂关联情境增加了解决该问题的难度。为此，本文首先构建了具有层级网络结构的风险因素识别框架，然后提出了一种考虑复杂关联情境的风险因素识别方法，将各专家针对系统关联和风险因素关联给出的语言评价信息转化为二元语义，并将决策试验与评价实验室（DEMATEL）法和Two-Additive Choquet（TAC）积分算子扩展到二元语义环境，进而实现各类关联的综合集成，从而确定风险因素的排序和归类，便于决策者研判风险根源、明晰风险因素本质。最后，以北京某样区为例验证了所提方法的潜在应用价值，并根据识别结果制定了针对性的风险防范策略，能够为相关部门联调联动开展风险防范提供决策支持。

Study on Identification Method for Risk Factors of Urban Typical Lifeline Operation Considering a Complex Interdependent Context

Heating, electricity and gas systems are representatives of urban typical lifeline with the characteristics of strong publicness, high risk and significant interdependence. Research on the problem of risk factor identification for the operation of urban typical lifeline has a very important significance for relevant departments to prevent risk. However, a complex interdependent context formed by multiple systems, multiple risk factors and several interdependences in reality increases the difficulty of solving the mentioned problem. Thus, a framework for risk factor identification with hierarchical network structure is firstly constructed in this paper. Then, a method for risk factor identification considering the complex interdependent context is proposed. In the method, linguistic assessment information on system interdependence and risk factor interdependence provided by experts are converted into two-tuple linguistics. Through extending the Decision Making Trial and Evaluation Laboratory (DEMATEL) method and Two-Additive Choquet (TAC) integral operator to two-tuple linguistic environment, several interdependences are integrated comprehensively. Furthermore, the rank and classification of risk factors are determined to help decision makers judge risk source and distinct the essences of risk factors. Finally, an example in Beijing is used to illustrate the potential application value of the proposed method, and the targeted measures for risk prevention are given according to the identification results. This study provides a decision support for relevant departments to conduct risk prevention collaboratively.