摘要
通过分析我国建筑施工安全生产的现状,在评价方法选择以及评价指标体系的建立方面,都在传统评价方法的基础上进行了改进,在BP神经网络的模型中引入了动量因子mc,从而克服了BP神经网络运算次数多,易陷入局部最优等缺点。利用改进的BP神经网络模型对建筑施工安全进行评价,找出施工中影响安全的主要因素,为建筑施工的危险辨识和制定相应控制措施提供依据,以指导人们有效地控制施工中的不安全因素,避免和减少施工中事故的发生。
Through analysis of the current situation of construction safety, the evaluation methods and index system of construction safety are improved based on traditional evaluation methods. A BP neural network model with momentum factor, mc, is proposed to overcome the weaknesses of BP neural network operation times and falling into local optimum. Main factors which affect construction safety are identified by using this improved BP neural network model. This model provides a reference for construction safety factors identification and risks controlling to minimize construction accidents.
出处
《工程管理学报》
2011年第6期629-632,共4页
Journal of Engineering Management
关键词
安全评价
改进的BP神经网络
危险因素
safety evaluation
BP newral network
construction safety factor
