Gaussian Model Approach for Identifying Car Engine Malfunctions Based on Embedded Sound System
Farhan Chandra Mahadika, Barlian Henryranu Prasetio · 2023
One important aspect for car owners is to maintain their vehicles in optimal and well-maintained condition. This includes tasks such as regular oil changes and system inspections. It is crucial for car owners to periodically maintain their vehicles through recommended manufacturer services. By relying solely on scheduled maintenance, it is possible that cars can provide indications of malfunction through engine sounds, which, if ignored without the owner's knowledge, can have negative consequences. Therefore, the identification of car engine issues using deep learning is essential for owners to receive early warnings of malfunctioning vehicles. Hence, the focus is on identifying car malfunctions based on engine sounds, utilizing MFCC and GMM for classifying normal and abnormal engine sounds. Consequently, a practical system is developed that can be implemented using the Raspberry Pi 3 B+ minicomputer.