基于随机滤波理论的剩余寿命预测方法研究

A Study on Residual Life Prediction Method Based on Stochastic Filtering Theory

状态预测是进行视情维修决策的关键和难点。针对工程实践中越来越多的非线性失效过程,利用随机滤波理论,基于当前时刻的状态监测历史信息,建立了含可变参数的状态预测模型。该模型克服了通用和两阶段状态预测模型的缺陷和不足,不仅能够动态地预测被监测设备在运行时的剩余寿命,而且能够准确灵活地反映出设备状态在失效过程中的变化,更接近于描述设备的真实运行过程。通过在MATLAB环境下对该模型进行仿真,验证了该模型的有效性。

State prediction is a critical and difficult problem in condition-based maintenance decisionmaking. Aiming to deal with more and more non-linear failure process in maintenance practice, a condition prediction model with floating parameters is established using the stochastic filtering theory given the condition monitoring history to date. The model that overcomes the deficiencies in the previous models ,can not only predict the residual life of the monitored asset dynamically, but also reflect the change of asset state in the failure process, which is closer to the underlying state than previous models. Based on MATLAB,the simulations demonstrate the validity of the proposed model.