Title :
Application of neural networks to detecting misfire in automotive engines
Author :
Ribbens, William B. ; Park, Jaehong ; Kim, Deeeun
Author_Institution :
Dept. of Electr. Eng. & Comput. Sci., Michigan Univ., Ann Arbor, MI, USA
Abstract :
This paper presents a novel application of neural networks to a vexing practical problem in the automotive industry. By government regulations, automobiles are required to be equipped with instrumentation to detect engine misfires and to alert the driver whenever the misfire rate has the potential to affect the health of emission control systems. A relevant model for the powertrain dynamics is developed in this paper as well as an explanation of the instrumentation. The basis for using a neural network to detect these misfires is explained and experimental system performance data (including error rates) are given. It is shown in this paper that the present method has the potential to meet the government mandated requirements
Keywords :
air pollution; air pollution control; automotive electronics; feedforward neural nets; internal combustion engines; multilayer perceptrons; automotive engines; automotive industry; emission control systems; engine misfires detection; error rates; experimental system; government regulations; instrumentation; misfire detection algorithm; misfire rate; neural networks; performance data; powertrain dynamics; Automobiles; Automotive engineering; Control systems; Electrical equipment industry; Engines; Government; Instruments; Mechanical power transmission; Neural networks; Power system modeling;
Conference_Titel :
Acoustics, Speech, and Signal Processing, 1994. ICASSP-94., 1994 IEEE International Conference on
Conference_Location :
Adelaide, SA
Print_ISBN :
0-7803-1775-0
DOI :
10.1109/ICASSP.1994.389586