<p>Precise localization is essential for accurately recognizing the current driving status of an autonomous vehicle and ensuring the safety of autonomous driving systems. Traditionally, global positioning system (GPS) based localization technologies have been widely used; however, in environments where GPS accuracy is compromised by external factors, fusion with other sensors such as inertial measurement unit (IMU) and odometry has been performed. Nevertheless, as mobility platforms evolve, vehicle behavior becomes increasingly complex, necessitating high-performance position estimation algorithms that account for both sensor and driving characteristics. However, the unique dynamics of four-wheel steering (4WS) vehicles, such as lateral movement without significant yaw-rate changes, pose challenges for conventional fusion algorithms in distinguishing intentional maneuvers from vehicle slip. To address this specific problem, this study proposes a hierarchical position estimation algorithm that decouples slip-aware velocity estimation from precise position tracking to resolve this ambiguity. The low-level module employs odometry and IMU data with filtering techniques to estimate velocity while accounting for factors such as vehicle and tire slippage. The high-level module then performs precise position estimation using the velocity estimated by the low-level module and GPS position data. The algorithm is validated using a platform equipped with a four-wheel steering system capable of both in-phase and out-of-phase steering, depending on the driving mode. The results reveal that the proposed hierarchical structure, with modules tailored to specific sensor characteristics and driving scenarios, offers improved robustness and better position estimation performance across a variety of conditions.</p>