Research on Civil Engineering Construction Safety Management Methods by Introducing Bayesian Network Modeling

Abstract

This paper aims to improve the effectiveness of civil engineering construction safety management (CSM)and ensure the safety of civil engineering construction. Firstly, an examination of contemporary CSM methods in the field of civil engineering in China is being undertaken. An innovative Civil Engineering Construction Safety Early Warning System (EWS) based on Bayesian Network (BN) principles addresses the unanticipated emergence of safety incidents during construction endeavors. Finally, a comprehensive evaluation model for civil engineering construction safety has been formulated, using the Bayesian framework to assess the efficacy of the previously mentioned Civil Engineering Construction Safety EWS based on BN. The incidence of safety-related accidents in the Chinese construction industry has consistently declined over the past decade, as the empirical findings indicate. Moreover, compared to international counterparts, the frequency of such incidents in China remains significantly below 600. Additionally, the Civil Engineering Construction Safety EWS created using BN consistently achieves a functional realization score exceeding 75 points, with the highest possible score reaching an impressive 93 points. Notably, through the utilization of the Civil Engineering Construction Safety evaluation model, which integrates Bayesian methodologies, it is discerned that machine tool quality and the placement of construction organization carry a substantial weight, each exceeding 0.3. Remarkably, the indicator of illegal operations is identified as having the highest risk level.