This thesis presents a novel approach to enhancing robotic autonomy through the design and simulation of a terrain-adaptive 4WD platform utilizing a low-cost MEMS Inertial Measurement Unit (IMU). Addressing the critical challenge of reliable navigation across diverse, unstructured environments, our work develops a robust framework for adaptive motion control and real-time terrain classification. Within Webots R2025a, the system first implemented adaptive control, dynamically adjusting wheel velocity based on IMU tilt and calculated slip, achieving a 28% reduction in average wheel slip on bumpy terrain. Subsequently, a real-time classification algorithm, fusing IMU vibration characteristics with slip data, demonstrated a 92% accuracy in distinguishing between simulated grass and asphalt surfaces. These capabilities enable the robot to intelligently modify its locomotion parameters, significantly improving stability and traction. Our core contribution lies in validating that accessible IMU technology, combined with sophisticated filtering and a rule-based decision architecture, provides an efficient and scalable solution for advanced robotic mobility. This research offers a strong foundation for developing more resilient autonomous systems vital for applications in challenging, real-world scenarios.