基于GA-BP的建筑施工安全实时预警模型  

Real-time Pre-alarm Model of Construction Safety Based on GA-BP

作　　者：刘洋 申玲[1] 陈东 LIU Yang;SHEN Ling;CHEN Dong(College of Civil Engineering, Nanjing Tech University, Nanjing 211816, China)

机构地区：[1]南京工业大学土木工程学院,江苏南京211816

出　　处：《土木工程与管理学报》2019年第2期167-172,185,共7页Journal of Civil Engineering and Management

基　　金：国家自然科学基金(71601095)

摘　　要：对建筑施工现场实时有效的预警是保证建筑施工安全进行的关键,目前建筑施工安全预警受到现场相关人员经验、项目现场环境动态变化等问题限制。对GA-BP法的研究表明,其可以避免预警过程复杂的参数估计过程并降低主观因素的影响。本文首先对施工现场安全信息需求进行系统的概括和划分,建立了安全信息需求指标体系,然后结合BLEMesh网络实时收集安全信息数据,最后引入遗传算法GA,利用遗传算法对BP神经网络进行优化,采用优化后的GA-BP网络对建筑施工现场进行安全预警,并通过事故实例分析说明该模型的适用性及优势。基于GA-BP的建筑施工安全实时预警模型可以有效实现建筑施工现场智能化、实时性、动态性的预警。The real-time and effective early warning of the construction site is the key to ensure the safety of construction. At present, the early warning of construction safety is limited by the restrictions on the experience of the field related personnel and the dynamic changes of the project site environment . The research of GA-BP method shows that it can avoid the complex parameter estimation process and reduce the influence of subjective factors in the early warning process. First, the safety information needs of the construction site are systematically summarized and classified, and the index system of safety information needs is established.Then, combined with BLEMesh network, safety information data are collected in real time. Finally, the genetic algorithm (GA) is introduced, and the BP neural network is optimized by genetic algorithm. The optimized GA-BP network is used for early warning. The applicability and advantages of the model are explained through accident case analysis . It is concluded that building safety real-time pre-alarm model based on GA-BP can effectively realize the intelligent, real-time and dynamic early warning of building site.

关 键 词：建筑施工 安全预警 人工神经网络 遗传算法 

分 类 号：TU714[建筑科学—建筑技术科学]