This thesis proposes a torque distribution strategy for skid-steering drive mobile robot locomotion in challenging terrains. The main objective is to promote effective torque conversion into motion. Employing an adaptive multi-objective optimization approach, the methodology derives equations correlating wheel torques to stable trajectories within the robot s workspace through a model-based genetic algorithm. Reactive torque control, facilitated by inertial sensors and torque feedback, enables automatic torque distribution control based on chassis orientation. This approach accommodates various individual objective functions, enabling adaptable behaviors for navigating beyond various obstacles. Experimental validation across diverse mobile robot platforms and configurations underscores the approach s efficacy, surpassing conventional techniques in terms of robustness and effectiveness. Real-time control implementation using a heuristic, correlating the robot s state with torque distribution, further enhances the maneuverability of skid-steer configurations in less-explored conditions, facilitating navigation through challenging terrains such as climbing steep slopes and negotiating small steps. By integrating optimization methodologies, the research contributes to advancing wheeled robot mobility, offering promising prospects for applications across industries.