Condition monitoring and fault diagnosis in a domestic car engine using a neural network

Abstract

The use of microelectronics in motor vehicle technology is well established, although a relatively recent innovation in comparison with the lengthy history of the motor car. In the automobile engine, the spark plug is in direct contact with the combustion processes and examination of the time varying voltage due to the spark ignition system at the spark plug provides a robust method of monitoring the engine. It has been shown that this waveform, subjected to neural network analysis, forms a mechanism which allows the monitoring of the engine for fault diagnosis or control. The technique potentially provides a convenient method of accurately sensing the /spl lambda/ of the engine. An investigation by the author and fellow researchers to demonstrate the feasibility of the technique has been described previously (R.J. Howlett et al., 1996); a summary of this work to date is presented.