轴承温度区间预测与稳定性分析

Temperature Interval Prediction and Stability Analysis for Bearings

通过温度检测装置测试了不同时刻下轴承的工作温度信息,获得2套轴承温度时间序列,每套包含50个数据。基于灰自助法对每组数据序列进行区间预测,利用前6个温度数据进行预报分析,后44个数据进行模型验证。预测结果表明,温度预测区间几乎包含所有试验值,误报率小,精度高。然后基于模糊集合理论进行轴承温度稳定性评估,挖掘出轴承温度变化趋势的本质特征及其性能退化迹象。

The temperature information of bearings under different times are measured through temperature detector,and the temperature time series of two sets of bearings are obtained with each set containing 50 data. The interval prediction is able to be realized for data series of each group based on grey bootstrap method. The former 6 temperature data is used to forecast and analyze,and the latter 44 data is adopted to verify model. The prediction results show that the prediction interval of temperature contains almost all the experimental values with a small rate of misinformation and a high precision. Then the temperature stability of bearings is evaluated based on fuzzy set theory,the performance degeneration signs and essential feature of variation trend of temperature of bearings are mined effectively.