Signal-based and Model-based Wheel Fault Detection of Omni-directional Vehicle with Mecanum Wheel

This paper uses an omni-directional four-wheel Mecanum vehicle to study the model-based and signal-based wheel fault behavior. In order to identify and detect the actual experiment where a wheel rotation is disabled manually. The two methods discussed approached this detection problem with or without a system model. Both methods use vibrational signal placed at different positions of the vehicle while the signal based identifies frequency shift at fault scenarios and the model-based method estimates the viscous effect coefficient through a simple on-line with the vibrational signals. The feasibility of the two methods through simulation and experiment as well as verify the proposed hypothesis.