Recently, controllers using the barrier Lyapunov function (BLF) are widely used to guarantee the tracking error (or state) constraint. A BLF can be defined only when the error is within a predefined boundary. Thus, overall stability can not be guaranteed when the error exceeds the predefined boundary owing to avoidable reasons such as external disturbance. Additionally, near the boundary, the singularity problem causes the peaking phenomenon in the control input. To overcome this problem, BLF based singularity-free nonlinear position control is designed for PMSMs. A state dependent boundary as a smooth function of the error is proposed to suppress the error when it exceeds the predefined boundary. When the error exceeds the predefined boundary, the proposed boundary increases according to the error. A nonlinear position controller is designed using the BLF and state dependent boundary. The tracking error always be controlled to be kept within the boundary by the position controller. When the tracking error exceeds the predefined boundary, the tracking error can return to the predefined boundary with the finite control input. The singularity problem can be solved by using the proposed boundary which is the smooth function. A disturbance observer is designed to estimate the disturbance. The performance of the proposed method is verified by simulations and experiments.
Motion sickness is a common condition experienced by drivers of skid-steer vehicles, primarily caused by zero crossing and oscillations in undamped systems. This study proposes an observer-based nonlinear control scheme to reduce transient oscillations and zero-crossing phenomena in skid-steer vehicles, thereby potentially alleviating motion sickness. Reducing transient oscillations and zero crossing in the transient response may alleviate motion sickness. A nonlinear damping controller is designed to improve transient response by reducing oscillations and zero-crossing. To design the controller, a reduced-order kinematic model based on coordinate transformation is developed. This transformation not only converts the system modeling into a controllable form but also enhances control performance. Modeling error is addressed by considering the distance between the center of the vehicle and the sensor location. Despite these improvements, model uncertainties and external disturbances remain, which may degrade control performance. To ensure robustness and estimate such disturbances, a high-order sliding mode observer (HOSMO) is incorporated. The effectiveness of the proposed method is validated through MATLAB/Simulink and TruckMaker simulations. From the simulation results, it was shown that the proposed method reduced the mean squared error of the tracking error to within 10 % compared to the state feedback controller with the HOSMO.