摘要
随着锚杆锚固技术在世界各地岩土工程中越来越广泛的应用,锚杆锚固工程质量问题时有发生,因此,对锚杆的检测工作就显得极其重要.以某工程为实例,利用MATLAB神经网络工具箱,通过选取训练样本和测试样本,建立神经网络模型,将频带能量作为特征向量,进行网络训练.训练结果理想,表明利用MATLAB神经网络工具箱进行锚杆缺陷的诊断是可取和准确的.
With the wider application in the geotechnical engineering of anchorage technique of anchors in the world,anchorage problems has been happening at all times,so the examination to the anchors seems to be very important.Taking an engineering as an example,neural netwoks can be built with the reasonable choice of training samples and testing ones,building the exact models and taking the energy of corresponding frequency bands as the properties of the networks,finally to be trained.The ideal results of the training show that taking advantage of neural network tools of of MATLAB,the diagnosis to defects of anchors is exact and reliable.
出处
《河北建筑工程学院学报》
CAS
2010年第1期116-118,共3页
Journal of Hebei Institute of Architecture and Civil Engineering
关键词
神经网络
锚杆缺陷
诊断
neural network
anchor defects
diagnosis
