Open Access

Luka Vasovic

• This thesis investigates the design and simulation-based evaluation of a yaw-rate control system using torque vectoring for an all-wheel-drive electric go-kart with independently driven wheels. The work is conducted within a purely model-based framework, employing MATLAB and Simulink to replicate the vehicle dynamics and control interactions in a controlled virtual environment. Torque vectoring describes the distribution of drive torque between individual wheels to influence the yaw moment of the vehicle. This enables targeted improvements in stability, responsiveness, and cornering performance beyond what is achievable with steering input alone. In this study, yaw rate is selected as the primary control variable. A reference yaw rate is derived from the linear bicycle model, and the control objective is to minimize the error between the reference and the simulated actual yaw rate by applying corrective yaw moments through differential torque allocation. A Proportional-Integral-Derivative (PID) controller is implemented to track the reference yaw rate. Gain tuning is performed using MATLAB’s PID Tuner, leveraging the system transfer function extracted from the simulation model. The complete control system is integrated into a modular Simulink model of the go-kart, which incorporates the 2-DOF bicycle model for lateral and yaw dynamics, as well as subsystems for maneuver generation and torque vectoring logic. Controller performance is assessed through standard vehicle dynamics test maneuvers, including ramp steer, step steer, and double lane change. For each maneuver, simulations are performed both with and without torque vectoring, allowing quantitative comparison of yaw rate tracking accuracy, stability, and trajectory. The results demonstrate that torque vectoring substantially improves yaw rate tracking and reduces the understeer tendency of the simulated vehicle, particularly during transient maneuvers. Even in a simplified small-scale vehicle model, the benefits of active yaw moment control are evident, underscoring the relevance of torque vectoring for enhancing dynamic performance in electric vehicles with independent wheel actuation. The modular simulation framework developed in this work also provides a foundation for future experimental validation on the physical research platform as well as further developments on the simulation model.