DocumentCode :
3475353
Title :
A PHM system for AEW radar based on AOPS-LSSVM prognostic algorithm and expert knowledge database
Author :
Xie Guang-jun ; Yan Shi-qiang ; Tang Zi-yue ; Rui Li
Author_Institution :
Air & Space Borne Early Warning & Surveillance Equip. Dept. of Air Force Radar Acad., Air Force Radar Acad., Wuhan, China
fYear :
2010
fDate :
12-14 Jan. 2010
Firstpage :
1
Lastpage :
6
Abstract :
In order to improve the ability of fault prognostic and the efficiency of fault diagnosis for certain AEW (airborne early warning) radar, in this paper, an APSO-LSSVM (adaptive particle swarm optimization- least squares support vector machine) fault prognostic algorithm and a fuzzy reasoning algorithm are presented, and an expert knowledge database is constructed too. Based on the APSO-LSSVM fault prognostic algorithm, fuzzy reasoning algorithm and expert knowledge database, a PHM (prognostic and health management) system is established for the AEW radar. The experiment shows that, because of using the APSO algorithm to adjust the parameters of LSSVM model, the APSO-LSSVM algorithm has a better fault prognostic ability; because of integrating the APSO-LSSVM algorithm with the fuzzy reasoning expert knowledge database, the PHM system not only can enhance the ability of health condition monitoring, but also can improve the efficiency of fault diagnosis and maintenance for the AEW radar. So, this PHM system can play a very important role in the AEW radar´s logistic support.
Keywords :
airborne radar; database management systems; expert systems; fault diagnosis; fuzzy reasoning; least squares approximations; particle swarm optimisation; radar computing; support vector machines; AEW radar; AOPS-LSSVM algorithm; PHM system; adaptive particle swarm optimization; airborne early warning radar; expert knowledge database; fault diagnosis; fault prognostic; fuzzy reasoning; least squares support vector machine; maintenance; prognostic and health management; Airborne radar; Condition monitoring; Databases; Fault diagnosis; Fuzzy reasoning; Least squares methods; Logistics; Particle swarm optimization; Prognostics and health management; Support vector machines;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Prognostics and Health Management Conference, 2010. PHM '10.
Conference_Location :
Macao
Print_ISBN :
978-1-4244-4756-5
Electronic_ISBN :
978-1-4244-4758-9
Type :
conf
DOI :
10.1109/PHM.2010.5413478
Filename :
5413478
Link To Document :
بازگشت