Search

Planning a safe and dynamic flyable trajectory for unmanned aerial vehicle (UAV) is the precondition for autonomous flight. Considering environmental complexity and flight safety under limited motor speed, we proposed a new time-segmented trajectory planning method, which can ensure motor speed within the allowable range and achieve safe flight. Firstly, the candidate trajectory for quadrotor UAV is expressed by high-order polynomials. Herein, polynomial is used as the expression of trajectory, and nonlinear programming (NLP) is used to calculate polynomial coefficient. Then, two strategies and an initial value calculation method are proposed to improve the efficiency of NLP. Furthermore, aiming at the problem that the initial trajectory of UAV motor cannot meet the current speed limit constraint after the motor speed drops, an adjustment strategy based on feasible trajectory is designed. The method presented in this paper not only enhances flexibility of the flight process, improves solving efficiency of the algorithm but also enables UAV to continue flying when the motor speed is limited. Finally, simulation analysis and comparison illustrate the effectiveness and superiority of the proposed method. [ABSTRACT FROM AUTHOR]

Trajectory tracking control of nonlinear system with disturbances has been extensively investigated. A novel fixed-time robust trajectory tracking control scheme for quadrotor unmanned aerial vehicle (UAV) with disturbances is proposed in this paper. First, the quadrotor system is split into position and attitude loops. With the purpose of restraining coupling disturbances on position loop forced by imperfect attitude tracking, a novel fixed-time unknown input observer (UIO) is devised to accurately estimate the disturbances, with the unique feature that does not need the upper bound of disturbances or the derivatives. Second, taking unmeasurable velocity and unknown external disturbances into account for attitude loop, a fixed-time extended state observer (ESO) is proposed to precisely observe them within a fixed time. Third, considering control accuracy and convergence rate, robust fixed-time controllers are, respectively, designed for position and attitude loops, and the convergence time of overall system is independent of system initial states. Then, a comprehensive stability proof for the two-loop system is presented with a combination of bi-limit homogeneity and Lyapunov technique. Finally, simulations and comparisons are performed to demonstrate the effectiveness of the designed control scheme. [ABSTRACT FROM AUTHOR]

Formation control is one of the key technologies for multiple unmanned aerial vehicles (UAVs). Compared with asymptotic or finite-time controllers, fixed-time controller can provide a guaranteed settling time, which does not depend on initial conditions and is an appealing property for controller design. Thus, robust fixed-time formation controller design for quadrotor UAVs under external disturbance and directed topology is investigated in this paper. A multi-variable super-twisting like integral sliding mode surface and a disturbance observer are respectively designed for position and attitude loops to guarantee robustness. Bi-limit homogeneity is utilized to design the whole closed-loop fixed-time controllers. By skillfully using bi-limit homogeneity technique and Lyapunov theory, the comprehensive stability of position and attitude loops is addressed. Finally, the multiple UAVs are utilized to track a pre-planned trajectory in 3D space and simulation results illustrate that the settling time can be reduced about 40% compared with finite-time controllers. [ABSTRACT FROM AUTHOR]

In this paper, finite-time fault-tolerant attitude tracking control is investigated for rigid spacecraft system with external disturbances, inertia uncertainties and actuator faults. A novel finite-time disturbance observer combined with a nonsingular terminal sliding mode controller is developed. Using an equivalent output error injection approach, a finite-time disturbance observer with simple structure is firstly designed to estimate lumped uncertainty. Then, to remove the requirement of prior knowledge about lumped uncertainty and reduce chattering, an adaptive finite-time disturbance observer is further proposed, and the estimations converge to the neighborhood of the true values. Based on the designed observer, a unified finite-time attitude controller is obtained automatically. Finally, both additive and multiplicative faults are considered for simulations and the results illustrate the great fault-tolerant capability of the proposed scheme. [ABSTRACT FROM AUTHOR]