Neural Adaptive Sliding Mode Control for Camera Positioner Quadrotor UAV

In the context of quadrotor UAV-based manipulation, it is important to achieve feedback control with faster convergence. However, achieving robust control and tuning of manipulator systems faces several challenges related to trajectory tracking due to parametric and model uncertainties. To address this, a Camera positioner-based quadrotor UAV (CPbQ UAV) is theoretically employed to track the velocity in this research. The study presents a control technique that combines a nonlinear asinh-based sliding surface with sliding mode control (asinhSMC) to meet the desired transient specifications. Asymptotic stability and tracking error convergence with the asinhSMC controller are analyzed using a quadratic Lyapunov function. To estimate the uncertain gravity functions in the rotational dynamics of the CPbQ UAV, an adaptive RBF neural network (RBFNN) is simulated. Performance characteristics such as settling time and overshoot of the proposed asinhSMC simulations are compared with PID and PIDSMC controllers for tracking desired step signals. Tracking simulations with the proposed control approach demonstrate its success for the CPbQ UAV, showing lower overshoot and settling time compared to PID and PIDSMC simulations, respectively.