A Knock Detection Method for a Marine Micro-Ignition Dual-Fuel Engine Based on Vibration Signals

Abstract A simple and unified index is proposed to achieve knock detection under various engine loads. Maximum amplitude vibration oscillation (MAVO) and maximum amplitude pressure oscillation (MAPO) were compared and were found to have no consistency. This means that MAVO cannot accurately reflect knocks inside the engine cylinder in the time domain. However, a knocking index built with MAVO can effectively detect engine knocking under various engine loads, which implies that some important information connected to the knock may be hidden within it. In this circumstance, a frequency domain analysis and a wavelet transform were conducted to study the energy changes of vibration signals during engine knocking. The energy proportion of the D1 frequency band during knocking increased drastically. Therefore, it was used to build a knocking judgment index, which builds the relationship between MAVO and MAPO. The judgment index has good applicability under different engine loads and a value greater than 0.5 can be used effectively for knock detection.