Research on scenario deduction and emergency decision-making evaluation for construction safety accidents

This study introduces an innovative “scenario-response†analysis framework for emergency management in construction safety accidents. Leveraging Bayesian Network (BN) technology, this framework systematically analyzes and evaluates the potential risks and developmental paths of such accidents. In-depth dissection of construction safety accident cases led to the creation of a comprehensive scenario knowledge meta-representation model. This model effectively captures both the micro-mechanisms and the macro-evolution of accidents. Furthermore, the study employs the entropy weight-TOPSIS model to scientifically evaluate emergency decision-making plans. This approach has resulted in the establishment of a comprehensive accident emergency response evaluation indicator system. It provides a quantitative basis for decision-making, aiding those managing construction safety accidents. The study’s empirical segment, focusing on a case analysis of a deep foundation pit collapse accident, illustrates key preventive measures. It highlights that timely reporting by supervisory personnel, enhanced supervision, and strict punishment of violations are crucial in preventing and controlling accident development in deep foundation pit construction. Additionally, the study validates the practical effectiveness and reliability of the proposed method. It offers valuable insights and guidance for both safety management practices and theoretical research within the construction industry.