Title :
MKLS-SVR based remaining useful life prediction for avionics
Author :
Yangming Guo; He Pei; Wu Hao;Jiaqi Zhang; Wang Zige; Jiang Xuefeng; Liu Junrui
Author_Institution :
School of Computer Science and Technology, Northwestern Polytechnical University, Xi´an 710072, China
fDate :
7/1/2015 12:00:00 AM
Abstract :
The avionic equipments are important parts of aircraft. Their failures take higher proportion in the total failure, which affect the performance of the whole system. A prediction model based on Multiple Kernel LS-SVR (MKLS-SVR) is proposed in this paper and used for remaining using life (RUL) prediction with a certain avionic device. The simulation results show that the MKLS-SVR has a higher accuracy comparing with the traditional LS-SVR, and it is a practical and effective electronic equipment RUL prediction method?
Keywords :
"Kernel","Predictive models","Mathematical model","Aerospace electronics","Data models","Time series analysis","Training"
Conference_Titel :
Electronic Measurement & Instruments (ICEMI), 2015 12th IEEE International Conference on
DOI :
10.1109/ICEMI.2015.7494257
