Novel Model Reference Adaptive Control Strategies for Direct Yaw Moment Control Applications

Path tracking control systems are vital for autonomous vehicles as they are responsible for the precise tracking of the trajectory decided by planning modules. However, the control of the later vehicle dynamics is challenging because of (i) the coupling between the longitudinal vehicle dynamics and the lateral dynamics (ii) tire nonlinearities and (iii) parameter uncertainties, such as the road friction coefficient, and the dependency of the system model from the longitudinal vehicle speed. From a control perspective, the control objective is to steer to zero the lateral position error of the vehicle with respect to the desired path while maintaining vehicle yaw stability. An effective way to improve vehicle stability is through the use of direct yaw moment control (DYC) systems. In such systems, the yaw moment adjustment is obtained from the difference in traction/braking forces between the left and right wheels, thus allowing precise control of the yaw rate. Hence, the aim of the project is to design control solutions that combine path tracking controllers with DYC systems. To deal with the aforementioned uncertainties, enhanced model reference adaptive control (EMRAC) solutions will be investigated during the project both for the control of the yaw rate via DYC and for path tracking/DYC. The EMRAC improves the tracking performance of classical MRAC by augmenting the control action with an adaptive integral control action and an adaptive switching control action. The integral action improves the rejection of tracking error in steady-state conditions while the switching action is used to suppress rapid bounded disturbances. Moreover, sigma-modification techniques guarantee the boundedness of the control gains also in presence of persistent disturbances. The control algorithms will be numerically validated in numerical software packages, e.g., MatLab/Simulink. (Keywords: EMRAC; Vehicle; Control; DYC; Path tracking)