Weighted bagging gaussion process regression to predict remaining useful life of electro-mechanical actuator

Abstract

Electro-Mechanical Actuator (EMA) is one of the key components of next generation aircraft. In order to ensure the safety of aircraft, it is critical to predict the remaining useful life (RUL) of EMA. And the RUL prediction can be implemented by utilizing Gaussian Process Regression (GPR). However, the GPR algorithm is extremely complex. Hence, a weighted bagging Gaussian Process Regression (WB-GPR) algorithm is presented in this article. To be specific, the significance of RUL prediction of EMA is analyzed, and the variable which can represent the degradation progress of EMA failure is selected. Then the framework to predict the RUL of EMA is realized, with the proposed WB-GPR. Finally the performance of RUL prediction based on WB-GPR is validated by utilizing the sensor data sets from National Aeronautics and Space Administration (NASA) Ames Research Center. Furthermore, the comparison of RUL prediction with GPR and bagging GPR has been achieved. Experimental results demonstrate that the WB-GPR is effective in the RUL prediction with low error rate and standard deviation.