Autonomous vehicle path-tracking and lateral stability depend critically on reliable steering actuator operation. However, steering systems are susceptible to bias faults from mechanical misalignment, friction, drivetrain asymmetry, and degradation. These faults distort commanded versus actual steering inputs, causing accumulated lateral and heading errors during high-speed driving. Actuator biases manifest as constant offsets, gradual drift, or intermittent activations, which complicate reliable diagnosis. This study presents an adaptive sliding mode observer-based active fault-tolerant control framework for real-time detection, estimation, and mitigation. An extended four-state lateral error model incorporating distance and heading errors captures the influence of steering bias on vehicle behavior and stability. Adaptive observer gain tuning addresses modeling uncertainties arising from speed variations, linearization residuals, and tire stiffness changes to ensure robust estimation under realistic driving conditions. The effectiveness of the proposed method is validated through high-speed double lane change simulations considering three representative bias scenarios: an initial constant bias, a gradually increasing drift bias, and an intermittent bias. Results demonstrate reliable bias estimation and significantly improved path-tracking accuracy compared to uncompensated cases. Operating without additional sensors, hardware redundancies, or controller switching, the framework is suitable for practical implementation in autonomous vehicle steering systems.

Accurate yaw angle information is essential for lateral stability and path tracking control of autonomous vehicles. Unlike actuator faults, GNSS-based yaw faults directly corrupt feedback states and can degrade vehicle dynamics, particularly at high speeds. In practice, GNSS yaw measurements are frequently affected by bias and drift caused by multipath effects, signal blockage, and degraded positioning conditions. This study proposes a control-oriented fault detection and reconstruction framework based on an adaptive sliding mode observer (ASMO) integrated with a lateral dynamic error model. The proposed ASMO employs an extended four-state error representation to maintain structural consistency between the observer and the path tracking controller. A multi-state adaptive injection structure adjusts the observer gain according to estimation errors, thereby reducing the oscillatory behavior of conventional sliding mode observers while maintaining robustness against nonlinear tire behavior and model uncertainty. In addition, a complementary fault detection strategy combining prediction-based feasible bounds and drift monitoring is developed to detect both abrupt bias faults and incipient drift faults. The reconstructed yaw signal is directly used in the control loop to support continuous fault-tolerant operation without controller switching. The proposed framework was validated through 100 km/h double lane change simulations in MATLAB/Simulink using a nonlinear vehicle model. Under a severe 5 ° GNSS yaw fault, the ASMO reduced the RMS lateral distance error from 0.3079 to 0.0401 m, corresponding to an 86.99% reduction compared with the uncompensated fault case. Although the RMS heading angle error slightly increased from 0.0427 to 0.0473 rad due to corrective steering behavior, the maximum lateral deviation was reduced from 0.4924 to 0.1761 m, corresponding to a 64.23% reduction, while maintaining a positive lane-boundary margin. These results demonstrate that the proposed ASMO enhances lateral fault-tolerant performance and safety-oriented functional safety in GNSS-challenged environments.