In order to effectively prevent the occurrence of dust explosion accidents in industrial enterprises and reducing the risk of dust explosion accidents, the article builds dust explosion risk assessment index system for industrial enterprises, taking Fault Tree Analysis (FTA) and Event Tree Analysis (ETA) as theoretical basis and establishes dust explosion risk Bow-Tie Analysis (BT) diagram model. Mapping BT model into Bayesian Networks (BN) reveals the coupling mechanism of dust explosion risk nodes and constructs a generic dust explosion risk assessment BN model, which provides a quantitative assessment of the probability of occurrence of dust explosion accidents. The causative mechanism of dust explosion accidents is investigated by learning risk inference from this model. It is also demonstrated that appropriate safety barrier measures can effectively reduce the probability of accidents, thereby changing the accident level and improving the safety and reliability of the system. Finally, the validity and applicability of the model is verified through empirical analyses. The research results contribute to the scientific prediction of the probability of dust explosion risk and reasoning learning, which provides a scientific basis for enterprises to formulate effective dust explosion safety measures as well as to realize the hierarchical control of dust explosion risk.

为了有效预防工业企业粉尘爆炸事故的发生，降低粉尘爆炸事故风险，文章采用故障树分析法（FTA）和事件树分析法（ETA）构建工业企业粉尘爆炸风险评估指标体系。为理论基础，建立粉尘爆炸风险蝴蝶结分析（BT）图模型。将BT模型映射到贝叶斯网络（BN）中，揭示了粉尘爆炸风险节点的耦合机制，构建了通用的粉尘爆炸风险评估BN模型，定量评估了粉尘爆炸事故发生的概率。通过学习该模型的风险推断，探讨粉尘爆炸事故的发生机制。事实证明，适当的安全屏障措施可以有效降低事故发生的概率，从而改变事故等级，提高系统的安全性和可靠性。最后通过实证分析验证了模型的有效性和适用性。研究成果有助于粉尘爆炸风险概率的科学预测和推理学习，为企业制定有效的粉尘爆炸安全措施、实现粉尘爆炸风险分级管控提供科学依据。