基于贝叶斯网络的主效应因素分析
Analysis of Main Effect Factors Based on Bayesian Network
摘要: 主效应因素分析是故障定位、道路交通事故分析、航空事故分析、风险分析、致因分析等领域关注和研究的重要课题，对于分析事故原因以及采取有效预防措施、降低事故发生率具有重要意义。现有基于贝叶斯网络的致因分析方法多数基于构建的贝叶斯网络结构对影响事故发生的因素进行单独的灵敏度分析，忽略因素之间的交互效应，从而得到片面的结论。本文提出一种基于贝叶斯网络的主效应因素分析方法，在充分利用因素之间的关联关系进行贝叶斯网络结构构建之后，分析各因素对结果因素的影响路径并进行联合影响度分析。实验证明，本文方法克服了简单割裂地对各因素进行单独分析的缺点，能够得到更加可靠全面的致因分析结论。
Abstract: Main effect factor analysis is an important topic in fault location, road traffic accident analysis, aviation accident analysis, risk analysis, cause analysis and other fields. It is of great significance to analyze the causes of accidents, take effective preventive measures and reduce the incidence of accidents. Most of the existing causal analysis methods based on Bayesian network are based on the constructed Bayesian network structure to conduct a separate sensitivity analysis of the factors that affect the occurrence of the accident, ignoring the interaction effect between the factors, so as to get a one-sided conclusion. In this paper, a main effect factor analysis method based on Bayesian network is proposed. After making full use of the correlation between factors to construct the Bayesian network structure, the influence path of each factor on the result factor is analyzed and the joint influence degree is analyzed. Experiments show that this method overcomes the shortcomings of simple split analysis of each factor, and can get more reliable and comprehensive causal analysis conclusions.
文章引用：仝小敏, 李国栋, 刘娜, 吉祥. 基于贝叶斯网络的主效应因素分析[J]. 计算机科学与应用, 2021, 11(5): 1236-1244. https://doi.org/10.12677/CSA.2021.115125

参考文献

[1] Long, Y.G., Wang, L.M., Duan, Z.Y. and Sun, M.H. (2019) Robust Structure Learning of Bayesian Network by Identi-fying Significant Dependencies. IEEE Access, 7, 116661-116675. [Google Scholar] [CrossRef] 
[2] 王辉, 张娟, 赵雅, 刘琨, 冯文峰. 一种新型贝叶斯模型的网络风险评估方法[J]. 小型微型计算机系统, 2020, 41(9): 1898-1904.
[3] 陈景文, 褚恩亮, 李英春, 杨亚莉. 基于小波包-贝叶斯的小电流接地故障区域定位方法[J]. 实验技术与管理, 2019, 36(8): 148-152.
[4] Heinrich, H.W. (2011) Industrial Accident Prevention: A Scientific Approach. Industrial & Labor Relations Review, 4, 609.
[5] 李俭川. 贝叶斯网络故障诊断与维修决策方法及应用研究[D]: [博士学位论文]. 长沙: 国防科学技术大学, 2002.
[6] 王波. 煤矿安全生产事故析因分析研究[D]: [硕士学位论文]. 徐州: 中国矿业大学, 2015.
[7] 张东, 聂百胜, 王龙康. 我国煤矿安全生产事故的致灾因素分析[J]. 中国安全生产科学技术, 2013, 9(5): 136-140.
[8] Foster, P. (2013) The Safety Journey: Using a Safety Maturity Model for Safety Planning and Assurance in the UK Coal Mining Industry. Minerals, 3, 59-72. [Google Scholar] [CrossRef] 
[9] 俞娉婷, 刘振元, 陈学广. 基于贝叶斯网络的一种事故分析模型[J]. 中国安全生产科学技术, 2006, 2(4): 45-50.
[10] 罗帆, 陈小佳, 顾必冲. 基于贝叶斯网络的航空灾害成因机理探析[J]. 北京航空航天大学学报, 2005, 31(8): 934-938.
[11] 秦小虎, 刘利, 张颖. 一种基于贝叶斯网络模型的交通事故预测方法[J]. 计算机仿真, 2005, 22(11): 230-232.
[12] 王发智. 基于贝叶斯网络的交通突发事件态势评估技术[D]: [硕士学位论文]. 大连: 大连理工大学, 2006.
[13] Marsh, W. and Bearfield, G. (2004) Using Bayesian Networks to Model Accident Causation in the UK Railway Industry. In: Spitzer, C., Schmocker, U. and Dang, V.N., Eds., Probabilistic Safety Assessment and Manage-ment, Springer, London, 3597-3602. [Google Scholar] [CrossRef] 
[14] Trucco, P., Cagno, E., Ruggeri, F. and Grade, O. (2008) A Bayesian Belief Network Modeling of Organizational Factors in Risk Analysis: A Case Study in Maritime Transportation. Engineering and System Safety, 93, 845-856. [Google Scholar] [CrossRef] 
[15] Friis-Hansen, P. and Simonsen, B.C. (2002) GRACAT: Software for Grounding and Collision Risk Analysis. Marine Structures, 15, 383-401. [Google Scholar] [CrossRef]