Acoustic emissions in the valuation of the combustion chamber pressure of an engine

Abstract

Internal combustion engines demand advanced monitoring methodologies to promote efficient operation; particularly, the combustion pressure plays a central role in the overall performance, which promotes the utilization of transducers that hinders. Therefore, the present study introduces an acoustic emission methodology that serves for indirect combustion pressure measurements. Accordingly, the compound methodology integrates the Hilbert transform and the complex cepstrum using neural networks to accomplish pressure signal reconstruction. Results demonstrated that the proposed methodology featured robust performance while estimating pressure signals as it mitigates the combined noise effect produced by variations in engine speed, engine load, and fuel type. Moreover, the reconstructed signal facilitated the determination of key performance parameters such as peak pressure, pressure timing, and effective mean pressure. Relative error amounted to less than 10%, which ratified the robustness of the indirect pressure measurements. In conclusion, acoustic signal techniques represent an adequate approach to estimate the combustion pressure at variable engine conditions.