退化数据驱动的设备剩余寿命在线预测

为在线预测单台服役设备的可用剩余寿命，提出一种融合先验退化数据和设备自身现场退化数据的剩余寿命预测方法。建立符合非线性Wiener过程描述的设备退化模型，利用先验数据采用极大似然法估计模型中的未知参数，使用贝叶斯方法融合新增的现场退化数据实时更新模型参数，进一步实现对设备实时剩余寿命评估。数值仿真和实例计算的结果表明，与固定参数法相比，该方法能够根据现场退化数据不断更新设备剩余寿命分布，进而更好地体现设备的个体差异，显著降低剩余寿命分布的不确定性。

Degradation data driven online prediction for equipment residual life

To predict available residual life of single service equipment, a prediction method of residual life which combines prior and current degradation data is proposed. The equipment degradation model is constructed, conforming to a nonlinear Wiener process. The unknown parameters are estimated by using the Maximum Likelihood Estimate（MLE） method. Parameters are updated by using the Bayesian method when new degradation data is available. After that, the real-time residual life is further evaluated. Numerical simulation and case study are conducted. Results indicate that the presented method can update the evaluation distribution of residual life by using new degradation data, reflect differences between individual equipment well and significantly reduce uncertainty of the residual life distribution, compared with the fixed parameter method.