AR模型应用于振动信号趋势预测的研究

振动信号的趋势预测是设备状态监测与故障诊断中的一个重要内容。论文提出采用时间序列ＡＲ模型进行趋势预测。将现场测得的非平稳振动序列通过ＡＲＩＭＡ模型和标准化处理，转化成标准正态平稳时间序列。模型参数估计使用了方法简单、参数估计无偏、精度高的最小二乘法。通过现场实测数据进行验证，计算结果表明ＡＲ模型能够很好地拟合振动信号时间序列并取得了一定的预测精度，可以达到预测要求。

Research on AR model applied to forecast trend of vibration signals

The trend forecasting of vibration signals is an important content of condition monitoring and fault diagnosis. AR model was used to forecast the trend of vibration signals. Through ARIMA model and standardization, the non stationary vibration series acquired in the field were transformed to stationary time series normally distributed. The simple, agonic and highly accurate least square method was used to estimate the model's parameters. Data acquisition in the field was used to verify this method. The experimental result indicates that the AR model can simulate time series of vibration signals quite well with good accuracy and meet the need of forecasting.