基于人工神经网络的建筑施工安全评价

Artificial Neural Network-Based Safety Evaluation for Construction Site

建筑施工现场的安全评价是一项复杂的系统工程。目前安全评价技术在建筑业的运用并不成熟,我国大多数建筑施工企业的安全管理只局限于对施工现场的检查和整改工作,而对整体安全性缺乏分析和有效监控。综合目前的安全评价技术,结合建筑施工的特点,确立建筑施工现场安全评价指标体系,并运用管理理论中的层次分析法(AHP)和模糊综合评价方法(Fuzzy),提出了适合建筑施工现场的人工神经网络(ANN)安全评价模型。详细论述了建筑施工安全评价方案以及具体实现的步骤,在结合AHP与Fuzzy综合评价法的基础上利用ANN进行训练与修正历史数据,为全面评价建筑施工安全状况提供了新的思路与方法。

Safety evaluation for construction site is a challenge problem.The technique for safety assessment in the construction industry is not well-developed.The safety management of most China construction companies concerns only construction site inspections and reforming.It lacks techniques for overall safety analysis and effective monitoring.With the characteristics of building construction considered,a safety evaluation index system is developed for building construction site by integrating the existing safety assessment techniques.With this index system,an artificial neural network（ANN）-based safety evaluation model is presented by combining analytical hierarchical process（AHP） and fuzzy theory.It is a new way for construction safety assessment.A case problem is used to show the detailed procedure for the application of the proposed method.