Combustion Monitoring Based on Engine Acoustic Emission Signal Processing

Acoustic emission contains information about several processes taking place within the engine. The combustion process could in fact be monitored by real time processing acoustic data, and also other features related to engine operation are contained in the very same signal (such as valve closing events, and both engine and turbocharger speed). The paper describes the development of real-time signal processing algorithms that could be integrated in the actual ECU software, in order to improve combustion diagnosis and control by extracting in-cylinder pressure rise rate information from the overall engine noise. In particular, the ability to effectively reconstruct in-cylinder pressure rise rate under all engine operating conditions would allow for a closed-loop combustion control system. Since the combustion acoustic emission signal to noise ratio may become particularly low, the paper shows that combustion quality may still be roughly recovered, thus providing information that can be used to monitor and diagnose abnormal combustion modes, such as knocking and misfiring (or partial burns).

Experimental tests have been carried out in a test-cell environment. Knocking and misfiring were externally induced, in order to evaluate the signal processing algorithms performance.