基于深度信念网络的造纸设备故障信号识别

Fault Signal Recognition of Papermaking Equipment Based on Deep Belief Network

针对现有造纸设备故障信号识别方法存在的精度性能低的问题,提出基于深度信念网络的造纸设备故障信号识别方法。根据造纸设备元件的故障机理,设置不同故障类型信号的识别标准。利用传感器设备采集造纸设备信号,通过中值滤波和中心化两个步骤,完成初始信号的预处理。利用深度信念网络,从时域和频域两个方面提取信号特征,通过特征匹配得出造纸设备的故障信号识别结果。经过性能测试实验得出结论:设计造纸设备故障信号识别方法的识别正确率指标均高于95%,即具有良好的识别精度性能。

Aiming at the problem of low accuracy and performance of existing fault signal identification methods for papermaking equipment, a fault signal identification method based on deep belief network is proposed. According to the fault mechanism of papermaking equipment components, the identification standards of different fault types are set up. The sensor equipment is used to collect the signal of papermaking equipment, and the initial signal is preprocessed through median filtering and centralization. Using deep belief network, signal features are extracted from time domain and frequency domain, and the fault signal recognition results of papermaking equipment are obtained through feature matching. Through the performance test experiment, it is concluded that the recognition accuracy index of the fault signal recognition method of papermaking equipment is higher than 95%, that is, it has good recognition accuracy performance.