This article develops an evolving type-2 quantum fuzzy neural network (eT2QFNN) control scheme for achieving trajectory tracking with unmanned aerial vehicles (UAVs). The proposed approach involves quantum membership functions, automatic rule growing process, and parameter adjustment learning scenario to deal with the problems of inadequacy, uncertainties, and noise in conventional control techniques. Furthermore, the proposed approach is operated in a parallel structure with the proportional derivative (PD) controller to compensate the transients in performance and learn the dynamic characteristics of the system. Besides, a sliding theory-based adaptive law is equipped with the control approach to compensate for the nonlinearity of the UAV. To assess the performance, numerical simulations and real-time experiments are carried for pitch and yaw axes control of two degrees of freedom (2DoF) helicopter test rig with the proposed approach. The simulations and experiments are aimed at achieving an offline path tracking with an objective to minimize the deviation error and improve the time response characteristics of the UAVs. The results depict the robustness of the proposed approach in terms of integral time absolute error for a helicopter following various trajectories. Note to Practitioners—This article addresses the problem of trajectory tracking and attitude control in a two-rotor UAV system. In practical application, there are multiple end users for an efficiently controlled UAV system. Generally, the trajectory tracking and attitude control are associated with the capability of UAVs to perform vertical take-off, landing, maneuver, and cyclic rotation as per the change in path. Furthermore, the control of UAVs for trajectory tracking and attitude provides a unified framework in efficiently following the desired path and maintaining the desired attitude. This has potential applications in the field of search and rescue operations, surveillance, military applications, environmental exploration, and aerial cinematography. Although trajectory tracking and attitude control problems have been studied a lot, the drawbacks due to uncertainties, immediate response to trajectory changes, and attitude settling are still open and challenging. This article proposed an eT2QFNN for a two rotor of UAV system with PD controller using automatic rule growing process. Sufficient trajectories are developed to make the UAV follow them under the proposed approach. Both simulation and real-time experiments were conducted and the results of the developed controller are compared with conventional approaches.
展开▼
