基于径向基神经网络的钢丝绳断丝定量检测及Matlab实现

Quantity Test of Steel Wire-rope LF Damage Base on RBF Neural Network and Matlab Realization

采用径向基神经网络对钢丝绳断丝损伤进行定量识别。首先将影响钢丝绳断丝损伤定量识别的6个主要因素作为输入参数,某截面的断丝数量作为输出参数建立起径向基神经网络模型,然后利用Matlab软件编写该网络的程序代码,通过有限的学习样本对网络进行训练,最后对测试样本进行识别。经过确认,测试结果较好地反应了钢丝绳的断丝损伤。

RBF neural network is adopted to test the LF damage quantity of steel wire - rope. It is necessary to define I/0 parameter before the network model is established. The six principal elements about LF damage are used for input parameter, and the output parameter is the number of LF damage. Then the procedure code is pmgramed with Matlab and the network should be trained with studying sample. After that, the tested sample is discerned by the network. After being confiemed ,the testing result is an exact match of the LF damage.