Acoustic based condition monitoring of a diesel engine using self-organising map networks

In this paper, the acoustic based condition monitoring of a diesel engine is investigated. Firstly, an experimental test rig is set up to simulate engine faults. Acoustic signals are measured from the test diesel engine under different conditions. To analyse the measured acoustic signals, a two-dimensional topological seif-organising map (SOM) network is employed in this paper to perform feature extraction. The extracted features are processed by both statistical and spectral methods. The results show that the extracted features are able to show the differences between the engine's normal and faulty conditions.