振动信号反映设备运行的状态,为了能够对设备的运行状态进行有效监控和预报,采用了频段振动烈度分析和ARIMA建模的方法。通过监控各个频段的振动烈度值,能够全面了解设备的状态。根据各个频段的变化趋势,选择处于上升阶段的频段进行ARIMA建模与状态预测。并讨论了如何处理现场数据的非平稳性以及模型类型判别方法,构建合适的时间序列模型。 The vibration signal reflects the running status of the equipment. In order to effectively monitor and predict the running status of the equipment, the methods of frequency vibration severity analysis and ARIMA modeling are adopted. By monitoring the vibration severity values ​​for each frequency band, you can get a complete picture of the status of your equipment. According to the changing trend of each frequency band, the ARIMA modeling and state prediction are selected in the frequency band in the rising phase. It also discusses how to deal with the non-stationarity of the scene data and the method of discriminating the model type, and build a suitable time series model.