Design and experimental validation of a filter-based allocation strategy for the longitudinal dynamics control in an autonomous motorcycle

Autonomous driving has expanded across various domains, with advancements in urban robo-Taxis, autonomous delivery vehicles, agricultural applications, and racing competitions. However, autonomous technology for two-wheeled vehicles remains under-explored, with limited literature and industrial demonstrations. Despite motorcycles already feature basic autonomous components like cruise control, traction control, and anti-lock braking systems (ABS), full autonomy demands integrated control of front braking pressure and rear regenerative torque. This work addresses the torque allocation problem inside the longitudinal dynamics control of an autonomous motorcycle, which is under development at Politecnico di Milano from a commercial fully-electric Energica Eva Ribelle. In particular, a complementary filter-based dynamic allocation strategy is proposed to distribute braking forces effectively, considering actuator limits. After a simulation-based tuning to handle the trade-off between braking distance and vertical load transfer variations on the wheels, an experimental validation has been carried out on a test track. Experimental results confirmed the effectiveness of the proposed approach.