基于模糊集和改进DS证据理论的危化品道路运输体系贝叶斯网络风险分析

针对危化品道路运输体系中影响因素信息的不确定性和专家知识推断贝叶斯网络中条件概率表时存在的主观性，提出了一种将模糊集和改进Dempster/Shafer证据理论（DS证据理论）、贝叶斯网络结合在一起的危化品道路运输体系的风险评价方法。根据危化品道路运输体系的影响因素建立了相应的风险评价体系，确定各层级的评价指标。将专家对各评价指标的评价意见代入高斯型隶属度函数构造隶属度矩阵，进行改进DS证据理论数据融合，得到多位专家评价意见融合后的概率值分布。利用贝叶斯网络的推理功能，得出危化品道路运输体系的风险等级和其中各评价指标的概率值分布，找出体系薄弱环节并进行分析。以沈阳某危险货物托运有限公司为例进行研究，结果表明该公司中危化品道路运输体系的风险值为0.611 8，风险等级为较危险（V₂），其中人员因素概率（60%）和管理因素概率（52%）所占权重较大，需要公司重点关注并加强管理。

Risk analysis of a Bayesian network for harmful chemicals road transportation systems based on fuzzy sets and improved Dempster/Shafer (DS) evidence theory

There is considerable uncertainty about the factors influencing road transportation systems of hazardous chemicals and the conditional probability table inferred from expert knowledge in a Bayesian network is also highly subjective. This paper proposes a risk assessment method for road transportation systems of hazardous chemicals which combines fuzzy sets and improved Dempster/Shafer (DS) evidence theory to give a Bayesian network for combined risk assessment of hazardous chemicals road transport systems. For each of the factors influencing hazardous chemicals road transportation systems, the corresponding risk assessment system was established to determine the evaluation indicators of each level. Expert evaluation of each evaluation indicator was then substituted into the Gaussian membership function to construct the membership grade matrix, and the data fusion of the improved DS evidence theory algorithm was carried out to obtain the basic probability distribution of the evaluation opinions after the fusion of multiple experts. The resulting evaluation indicators allowed the weak links in the system to be identified and analyzed. Taking a hazardous goods consignment company in Shenyang as an example, the results show that the risk in transporting hazardous chemicals is relatively high (V₂), with the probability of human factors being responsible of 60% and the probability of management factors being 52%. Both of these have a large weight, which requires the company to focus on strengthening its management structure.