按BP神经网络的基本原理和算法,确定了振动筛的BP神经网络结构,用振动筛运行状态的特征量作为BP神经网络的输入,运用Matlab神经网络工具箱对该网络进行训练,得到了用于诊断的BP神经网络模型。实验结果表明,运用神经网络方法能较为准确诊断振动筛故障。
According to the principle and algorithm of BP neural network, the structure of vibrating sieve can be determined. The BP neural network model for fault diagnosing is obtained by taking the characteristic variables of operating state of vibrating sieve as the input of neural network and using the Matlab neural network toolbox for the network training. The experimental result shows that the method can accurately diagnose the fault of vibrating sieve.