Characterization of Braking and Clutching Events of a Vehicle Through OBD II Signals

This work presents an algorithm capable of identifying two habitual driving maneuvers such as braking to decrease speed and disengaging to produce a gear change in the vehicle by studying the PID’s (Identification Parameters) signals from the electronic control unit. These signals are acquired through the OBD II connector through a data logger device capable of storing information during the engine operation. The obtained data is prost-processed using K-mean unsupervised learning algorithm. The algorithm is capable of identifying braking and clutch events in addition to classifying whether the vehicle has a motorized or mechanical body acceleration system. With the used algorithm, it is possible to determine the pilot’s driving style and the most frequently used change during a route. For this research, four vehicles from different years and manufacturers were used to verify the functionality of the algorithm.