基于贝叶斯网络的装配式建筑安全风险研究

Research on Safety Risk of Prefabricated Buildings Based on Bayesian Network

为提高EPC总承包商对装配式建筑项目安全风险的识别和分析能力,通过贝叶斯网络算法对安全风险因素进行定量分析和评估。先通过安全数据采集建立评价指标,采用熵权法计算各安全因素的权值,结合风险指标的发生概率与影响程度,构建风险矩阵,再通过构建贝叶斯网络模型,获取各层级风险的概率分布并逆向推理得出影响安全风险的最关键因素及其概率分布。研究成果能够有效提高EPC总承包商对装配式建筑项目安全风险的识别和分析能力,为装配式建筑项目的安全管理提供参考。

Aims to improve EPC general contractor’s ability to identify and analyze security risks of prefabricated construction projects,and to quantitatively analyze and evaluate security risk factors through Bayesian network algorithm.Firstly,the evaluation index is established through the collection of safety data,the weight value of each safety factor is calculated by entropy weight method,and the risk matrix is constructed by combining the occurrence probability and influence degree of risk indicators.Then,the probability distribution of risks at all levels is obtained by constructing Bayesian network model,and the most critical factors affecting safety risks and their probability distribution are obtained by reverse reasoning.The research results of this paper can effectively improve the EPC general contractor’sability to identify and analyze the safety risks of prefabricated construction projects,and provide reference for the safety management of prefabricated construction projects.