Title :
A forecasting method for remaining useful life based on EMD denoising algorithm and Wiener-process
Author :
Zhang Jian-xun ; Hu Chang-Hua ; Zhou Zhi-jie ; Zhang Zheng-xin ; Zhang Jia
Author_Institution :
Fac. 302, Second Artillery Eng. Univ., Xi´an, China
Abstract :
This paper presents a modified EMD method by improving the traditional polynomial prediction method of restraining the end effects, and then presents a denoising method based on EMD and soft threshold method. This paper builds Wiener-process-based degradation model by the denoised data, and presents a forecasting method for remaining useful life based on EMD denoising algorithm and Wiener process. The experimental analysis result demonstrates that the method can decrease modeling error and increase the remaining useful life estimation´s degree of accuracy.
Keywords :
Hilbert transforms; signal denoising; stochastic processes; Wiener-process-based degradation model; forecasting method; modified EMD denoising algorithm; soft threshold method; traditional polynomial prediction method; useful life; Algorithm design and analysis; Degradation; Educational institutions; Electronic mail; Forecasting; Noise reduction; Prediction algorithms; Degradation Failure; EMD; Residual Lifetime; Weiner Process;
Conference_Titel :
Control Conference (CCC), 2013 32nd Chinese