DocumentCode
3231036
Title
Linguistic model for engine power loss
Author
Grantner, Janos ; Bazuin, Bradley ; Fajardo, Claudia ; Hathaway, Richard ; Al-shawawreh, Jumana ; Dong, Lixin ; Castanier, Matthew P. ; Hussain, Shiraz
Author_Institution
Dept. of Electr. & Comput. Eng., WMU, Kalamazoo, MI, USA
fYear
2013
fDate
16-19 April 2013
Firstpage
80
Lastpage
86
Abstract
Army ground vehicles often operate in extremely severe environmental and battlefield conditions. Condition Based Maintenance (CBM) allows maintenance to be performed based on evidence of need provided by reliability modeling and/or other enabling technologies, thus reducing maintenance costs and increasing vehicle availability. A Takagi-Sugeno fuzzy model is developed to diagnose the loss of engine power of light trucks. Baseline data are acquired through engine performance measurements. The Adaptive Neuro-Fuzzy (ANFIS) training method is used to extract the fuzzy rules. To improve the quality of the model a combination of the least-square error and the backpropagation gradient descent methods is implemented to minimize the errors.
Keywords
diesel engines; fuzzy control; fuzzy neural nets; learning (artificial intelligence); maintenance engineering; military vehicles; reliability; ANFIS training method; CBM; Takagi-Sugeno fuzzy model; adaptive neuro-fuzzy training method; army ground vehicles; backpropagation gradient descent methods; battlefield conditions; condition based maintenance; engine performance measurements; engine power loss; extremely severe environmental conditions; fuzzy rules; least-square error; light trucks; linguistic model; maintenance cost reduction; reliability modeling; Engines; Loss measurement; Mathematical model; Power generation; Power measurement; Sensors; Vehicles; Condition Based Maintenance; engine power loss; fuzzy model; intelligent diagnostics;
fLanguage
English
Publisher
ieee
Conference_Titel
Computational Intelligence in Vehicles and Transportation Systems (CIVTS), 2013 IEEE Symposium on
Conference_Location
Singapore
Type
conf
DOI
10.1109/CIVTS.2013.6612293
Filename
6612293
Link To Document