Model-free Intelligent Control for Antilock Braking Systems on Rough Terrain

Advancements have been made in the field of vehicle dynamics, improving handling and safety through control systems such as Antilock Braking System (ABS). An ABS enhances braking performance steerability a under severe conditions by preventing wheel lockup. However, its degrades on rough terrain resulting an increased lockup stopping distance compared to without. This is largely result noisy measurements, un-modelled dynamics that occur vertical torsional excitation experienced over terrain. Therefore, it proposed model-free intelligent technique, which may adapt these be used overcome this problem. The Double Deep Q-learning (DDQN) technique conjunction with Temporal Convolutional Network (TCN) algorithm, straight line simulations are performed using single tyre model, characteristics approximated LuGre model. modelled after measured Belgian paving normal forces at contact patch FTire ADAMS. Comparisons drawn against Bosch results show approach achieves lateral stability whilst terrain, without deteriorating distance.