基于灰关联与神经网络的混凝土中钢筋锈蚀预测  

PREDICTION FOR REINFORCEMENT CORROSION IN CONCRETE BASED ON GRAY CORRELATION AND NEURAL NETWORK

作　　者：安明喆[1] 马凌飞[1] 闫志刚[1] 

机构地区：[1]北京交通大学,北京100044

出　　处：《建筑技术》2009年第12期1109-1111,共3页Architecture Technology

摘　　要：基于灰关联识别方法,解析各钢筋锈蚀因子对钢筋锈蚀量的影响程度;将BP神经网络的学习样本进行归一化及乘以灰色关联权系数处理,然后BP神经网络对处理后的样本进行学习,以对钢筋锈蚀量进行预测,结果证明该方法对钢筋锈蚀量能进行良好的预测。This paper applied gray correlation analysis to analyze the effects of the influencing factors to the reinforcement corrosion. First, normalize the learning sample of the BP neural network and multiply the coefficient of weight; then BP neural network study the sample treated in order to prediction the corrosive degree. The results proved that this method cans prediction well.

关 键 词：灰关联 BP神经网络 钢筋锈蚀量 预测 

分 类 号：TU511[建筑科学—建筑技术科学]