基于模糊证据推理的系统风险分析与评价

System risk analysis and evaluation approach based on fuzzy evidential reasoning

针对系统风险分析与评价中面临的模糊不确定风险信息建模问题,以及风险因素信息融合问题.在模糊集理论与证据理论研究的基础上.采用模糊证据推理技术,提出了基于模糊证据推理的系统风险分析与评价方法的具体算法与处理操作流程.首先通过故障模式及影响分析进行风险识别,采用模糊信度结构模型建立风险不确定性的统一描述框架,然后通过数据预处理算法、模糊证据推理信息融合算法以及模糊风险排序方法完成风险评价结果计算;文末以某航天发射场火箭推进剂加注系统风险分析为例,验证了方法的可行性与有效性.该方法定量化分析系统部件的风险程度,为后续风险应对措施的制定提供辅助决策支持.

In order to deal with problems of risk modeling under fuzzy uncertainty and risk information fusion, the risk analysis and evaluation approach based on fuzzy evidential reasoning and its solution procedure are proposed on the basis of fuzzy set theory and Dempster-Shafer theory of evidence. Firstly, failure mode and effects analysis is used for identifying risk factors and fuzzy belief structure is for building the uniform risk model. Secondly, the original qualitative and quantitative risk data is transformed into fuzzy belief structures using the pretreatment algorithm. Then, risk information fusion is completed using the fuzzy evidential reasoning and fuzzy intersection match algorithm, and risk evaluation result is ordered by the fuzzy risk ranking method. Finally, a risk analysis example of rocket propellant adding system in space launch site is examined to illustrate and show the applicability of the approach. The proposed approach may successfully analyze the system risk degree and provide decision making support for risk response.