分享兴趣，传播快乐，

增长见闻，留下美好！

亲爱的您，这里是LearningYard学苑！

今天小编给大家分享的是论文精读系列19。欢迎您的用心访问！

本次分享的硕士论文题目是《基于贝叶斯网络的农超对接供应链风险预警模型研究》，本期将学习第五章内容，一起来看看吧！

Share interests, spread happiness,

Grow what you see and leave behind!

Dear you, this is LearningYard Academy!

Today's Xiaobian to share with you is the paper intensive reading series 19. Welcome to your attentive visit!

The title of the doctoral thesis shared this time is "Research on the Risk Early Warning Model of Agricultural Super Docking Supply Chain Based on Bayesian Network", and this issue will learn the content of the 5th chapter, let's take a look!

本章内容主要为“农超对接”供应链风险预警模型的仿真分析。

The content of this chapter is mainly the simulation and analysis of the risk early warning model of the "agricultural super docking" supply chain.

目前主流的贝叶斯网络分析软件包括GeNIe、Netica、Matlab等。Netica软件不仅支持图形和编程接口，还具有GIS和2D分析等功能，具有简单、可靠、高效的特点，因此本文选用Netica软件进行“农超对接”供应链风险预警模型的仿真分析。

At present, the mainstream Bayesian network analysis software includes GeNIe, Netica, Matlab, etc. Netica software not only supports graphics and programming interfaces, but also has GIS and 2D analysis functions, with simple, reliable and efficient characteristics, so this paper uses Netica software for simulation analysis of the "agricultural super docking" supply chain risk warning model.

首先，根据预警灯号区间计算获取各个风险预警指标（包括一级指标、二级指标和总体指标）的风险值与灯号状态，对当前各项风险指标的风险水平进行初步评估分析，得到仿真预警模型的贝叶斯网络、二级风险指标、一级风险指标计算结果如下所示：

Firstly, according to the calculation of the early warning light range, the risk value and lamp status of each risk early warning index (including the first-level indicator, the second-level indicator and the overall indicator) are obtained, and the risk level of the current risk indicators is preliminarily evaluated and analyzed, and the calculation results of the Bayesian network, the second-level risk indicator and the first-level risk index of the simulation early warning model are as follows:

根据二级风险指标能够清晰看出个指标预警灯的颜色，其中红色的有2个，说明风险较高；黄灯有5个，说明存在一定风险；其他为绿灯或者蓝灯，表明风险水平较低或者极低。

According to the secondary risk indicators, the color of the indicator warning lights can be clearly seen, of which 2 are red, indicating that the risk is higher; There are 5 yellow lights, indicating that there is a certain risk; Others are green or blue, indicating a low or very low level of risk.

根据一级风险指标表格，企业管理风险、生产管理风险、对接合作风险预警灯号为黄灯，表示存在一定风险，需要注意；技术风险和外部风险为绿灯，表示风险较低。从整体风险表格种可以看出整体风险值为0.312，预警灯为黄灯，处于中等风险水平，需要引起关注。

According to the first-level risk indicator table, the early warning light number of enterprise management risk, production management risk and docking cooperation risk is yellow light, indicating that there are certain risks and need to be paid attention to; Technical risks and external risks are green lights, indicating lower risk. From the overall risk table, it can be seen that the overall risk value is 0.312, and the warning light is yellow, which is at a medium risk level and needs to be paid attention to.

然后，作者运用两个案例进行了正向仿真推理。正向推理是指根据接收到的新信息，基于对新信息的推理分析，对贝叶斯网络中相应风险因素节点的参数进行适当调整，从而借助贝叶斯网络的动态传导特性将信息传递至相关关键风险节点，根据关键风险节点的变化来分析风险状态，是定性的先验知识向定量的风险预测数据的有效转换。

The authors then used two cases to perform forward simulation inferences. Forward inference refers to the appropriate adjustment of the parameters of the corresponding risk factor nodes in the Bayesian network based on the inference and analysis of the new information received, so as to transmit the information to the relevant key risk nodes with the help of the dynamic conduction characteristics of the Bayesian network, and analyze the risk state according to the changes of the key risk nodes, which is an effective conversion of qualitative prior knowledge to quantitative risk prediction data.

接着，作者对整体风险指标进行了逆向推理。逆向推理又称诊断推理，是指对关键风险节点的参数进行调整，根据下级指标节点的变化来识别出关键影响因子。逆向推理主要用于识别关键风险因素，以及出现问题后的诊断分析。根据所构建的预警模型从整体、生产管理、企业管理、对接合作、技术以及外部 6 个方面进行逆向推理计算并排序。

The authors then reverse-reasoned about the overall risk indicator. Reverse inference, also known as diagnostic reasoning, refers to the adjustment of the parameters of key risk nodes and the identification of key influencing factors according to the changes of subordinate indicator nodes. Reverse inference is mainly used to identify key risk factors and to analyze diagnostics after problems occur. According to the constructed early warning model, reverse inference is calculated and sorted from six aspects: overall, production management, enterprise management, docking cooperation, technology and external.

总结：本章首先结合两个实际案例进行正向推理仿真预测分析，预测了 “农超对接”供应链各方面风险状态。其次通过逆向推理仿真进行风险诊断分析，识别了“农超对接”供应链各方面风险的关键风险因素，为相关管理人员提供资源优化配置的决策参考。

Summary: This chapter first combines two practical cases to conduct forward inference simulation and prediction analysis, and predicts the risk status of all aspects of the "agricultural super docking" supply chain. Secondly, the risk diagnosis and analysis is carried out through reverse inference simulation, and the key risk factors of all aspects of the "agricultural super docking" supply chain are identified, and the decision-making reference for the optimal allocation of resources is provided for relevant management personnel.

本期的分享就到这里，如果您对今天的文章有独特的想法，欢迎给我们留言，让我们相约明天，祝您今天过得开心快乐！

The sharing of this issue is here, if you have a unique idea for today's article, welcome to leave us a message, let us meet tomorrow, I wish you a happy and happy day!

翻译参考来源：Google翻译。

内容参考来源：

[1]谢乐.基于贝叶斯网络的农超对接供应链风险预警模型研究.

部分资料图片来源于参考文献，其余内容由LearningYard学苑原创，仅代表作者个人观点，如有侵权请沟通。

文字、排版｜圈儿

审核｜Tian