CAJ | 학술논문

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。
이용BP신경망락대강사승신호적특정량화단사수진행훈련，선택표준적3층망락장실험수거진행귀일화후변환은층절점수반복다차훈련，학정은층절점수，장효과비교호적망락작위실제단사검측시조용적망락，통과VC＋＋어언편정，실현강사승검측적지능화。
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.