基于BP神经网络的钢丝绳损伤定量识别

Quantitative identification of steel rope damage based on BP neural network

利用BP神经网络对钢丝绳信号的特征量和断丝数进行训练,选择标准的3层网络将实验数据进行归一化后变换隐层节点数反复多次训练,确定隐层节点数,将效果比较好的网络作为实际断丝检测时调用的网络,通过VC++语言编程,实现钢丝绳检测的智能化。

BP neural network is used to exercise the characteristic quantity of steel rope signals and fracture of wires.The standard 3-layer network is selected to normalize the test data,followed by multiple exercises by changing the quantity of hidden nodes.The quantity of hidden nodes is determined,with the most effective network taken as the one to be called when detecting actual fracture of wires.VC＋＋language programming is used to realize intelligent detection for steel ropes.