基于PSO-BP网络预测模型的造纸机压榨辊振动特性建模分析  

Modeling and Analysis of Vibration Characteristics of Paper Machine Press Roll Based on PSO-BP Network Prediction Model

作　　者：雷枫 LEI Feng(XIAN HaiTang Vocational College,Xi'an,710038,China)

机构地区：[1]西安海棠职业学院,陕西西安710038

出　　处：《造纸科学与技术》2022年第5期63-68,共6页Paper Science & Technology

摘　　要：了解造纸机压榨辊振动特性对于判断其工作状态具有十分重要的作用。在此背景下,基于PSO-BP网络预测模型分析造纸机压榨辊振动特性。利用振动传感器采集振动信号并通过小波阈值法实现振动信号去噪处理;提取振动信号中的时域特征(有效值、峭度值)和频域特征(功率谱、倒频谱)。以特征为输入,以振动特性定量值为输出,训练PSO-BP网络,完成预测模型的构建,利用预测模型预测造纸机压榨辊未来一段时间的振动特性,分析其工作状态是否正常。结果表明:未来一周内造纸机压榨辊振动特性定量值一直曲折上升并在5~6天之间振动特性定量值超过边界值,从健康区间进入异常区间。Understanding the vibration characteristics of the press roll of paper machine is very important for judging its working state. Under this background, the paper machine press roll vibration characteristics are analyzed based on PSO-BP network prediction model. Vibration signals are collected by vibration sensors and denoised by wavelet threshold method. The time-domain features(RMS, kurtosis) and frequency-domain features(power spectrum, cepstrum) of vibration signals are extracted. With the characteristics as the input and the quantitative value of vibration characteristics as the output, the PSO-BP network is trained to complete the construction of the prediction model. The prediction model is used to predict the vibration characteristics of the paper machine press roll in the future, and analyze whether its working state is normal. The results show that in the next week, the quantitative value of the vibration characteristics of the press roll of the paper machine has been zigzagging and rising, and the quantitative value of the vibration characteristics exceeds the boundary value between 5 and 6 days, from the healthy range to the abnormal range.

关 键 词：PSO-BP网络预测模型 造纸机压榨辊 振动特征 振动特性 

分 类 号：TP223.63[自动化与计算机技术—检测技术与自动化装置]