Two-wheeler tyre pressure monitoring through K-nearest neighbours algorithm trained using wheel hub vibrations acquired using ADXL335 accelerometer

Maintaining optimal tyre pressure enhances the performance of a vehicle in many ways. Tyre pressure monitoring system (TPMS) provides a safety feature that shows an alert when the car's tyre pressure drops below the recommended levels. In this paper, the TPMS is built through training of the KNN algorithm based on wheel hub vibrations. An inexpensive system was developed by interfacing the ADXL335 accelerometer with Arduino for collecting real-time data. In order to process the raw data suitable conditioning was undertaken. The initial judgment of wkNNheel hub vibrations was carried out statistically. The features reflecting the relevant statistical judgment of tyre pressure conditions were selected and training of the KNN algorithm was initiated. Perfectly filled, partially filled and unhealthy tyre conditions were considered while acquiring the wheel hub vibrations and classification was achieved.