Hybrid vehicle engine misfire detection using Piezo-film Sensors and analysing with Z-freq / Nor Azazi Ngatiman and Mohd Zaki Nuawi

This paper explores the application of standard deviation and kurtosis to formulate a new statistical method for hybrid electric vehicle (HEV) engine misfire detection named as Z-freq. Existing engine management system is unable to evaluate the performance of each individual piston of the internal combustion engine. Eventually, it will affect the whole hybrid system especially charging or cruising. The principal objective of this research is to diagnose the engine wall vibration and identify the combustion strokes. Furthermore, the primary advantages of this technique are the simple, accurate and fast computation of the algorithm such it can be processed rapidly using a low-cost
controller. This investigation involves two phases of the procedure including signal measurement using data acquisition equipment as well as signal analysis using the MatLab. At the first phase, a piezoelectric-based sensor
called piezo-film is attached to each engine wall, run the hybrid engine at different speeds and also run with misfire spark plug and record the signal measured. While the second phase is signal measurement using data acquisition, filtering and analysis using Z-freq method. The advantage of the
proposed method isthis technique able to identify different combustion strokes.
This technique is also can be applied for engine technical expert who requires simple, fast yet accurate results. These findings provide a solid evidence base on pattern representation which easier to the investigator to categorise the measured parameters.