Misfire Detection in Automotive Engines Using a Smartphone through Wavelet and Chaos Analysis

Area of Science:

• Automotive Engineering
• Mechanical Engineering
• Signal Processing

Background:

• Engine failures can compromise vehicle motion and passenger safety.
• Current vehicle failure detection systems lack real-time driver feedback.
• Engine misfire, caused by weak or absent spark, is a common and critical failure.

Purpose of the Study:

• To compare two distinct algorithms for identifying engine misfire failures.
• To evaluate the effectiveness of Wavelet Multiresolution Analysis (WMA) and Signal Analysis based on Chaos using the density of maxima (SAC-DM).
• To validate these methods on a working automotive vehicle under various failure conditions.

Main Methods:

• Experimental validation of WMA and SAC-DM algorithms on a vehicle.
• Testing scenarios included a healthy engine, single-piston misfire, and double-piston misfire.
• Analysis of engine signals to detect anomalies indicative of misfire.

Main Results:

• Both WMA and SAC-DM achieved 100% accuracy in diagnosing engine misfire across all tested conditions.
• The SAC-DM method demonstrated a capability for faster failure diagnosis compared to WMA.
• Experimental tests confirmed the reliability of both techniques in real-world automotive applications.

Conclusions:

• WMA and SAC-DM are highly effective methods for real-time engine misfire detection.
• SAC-DM offers a potential advantage due to its reduced time-to-diagnosis.
• These algorithms can significantly enhance automotive failure characterization and diagnostics.