UAV Attitude Control Research Articles

Large-Scale Solar-Powered UAV Attitude Control Using Deep Reinforcement Learning in Hardware-in-Loop Verification

Large-scale solar-powered unmanned aerial vehicles possess the capacity to perform long-term missions at different altitudes from near-ground to near-space, and the huge spatial span brings strict disciplines for its attitude control such as aerodynamic nonlinearity and environmental disturbances. The design efficiency and control performance are limited by the gain scheduling of linear methods in a way, which are widely used on such aircraft at present. So far, deep reinforcement learning has been demonstrated to be a promising approach for training attitude controllers for small unmanned aircraft. In this work, a low-level attitude control method based on deep reinforcement learning is proposed for solar-powered unmanned aerial vehicles, which is able to interact with high-fidelity nonlinear systems to discover optimal control laws and can receive and track the target attitude input with an arbitrary high-level control module. Considering the risks of field flight experiments, a hardware-in-loop simulation platform is established that connects the on-board avionics stack with the neural network controller trained in a digital environment. Through flight missions under different altitudes and parameter perturbation, the results show that the controller without re-training has comparable performance with the traditional PID controller, even despite physical delays and mechanical backlash.

Correction: Li et al. Attitude Control of UAVs with Search Optimization and Disturbance Rejection Strategies. Mathematics 2023, 11, 3794

In the original publication [...]

Attitude Control of UAVs with Search Optimization and Disturbance Rejection Strategies

This study aims to achieve rapid and stable control of quadrotor unmanned aerial vehicles’ (UAVs) attitude by using an Active Disturbance Rejection Control (ADRC) controller. Addressing the challenge of numerous and complex ADRC parameters, optimization algorithms are employed for parameter tuning. This paper draws on the group mechanism of the Ant Colony Optimization (ACO) algorithm and innovatively introduces population search into the Beetle Antennae Search (BAS) algorithm. The refined algorithm is then applied to tune the ADRC parameters, reducing complexity and human intervention while enhancing intelligence and efficiency. The advanced optimization algorithm exhibits an exceptional global optimization capacity, convergence speed, and stability. Ultimately, flight simulation and experimental results suggest that the optimized ADRC controller demonstrates superior control and antidisturbance capabilities.

Full-State-Constrained Adaptive Control for a Class of UAVs Suffering from Coupled Uncertainties Using the HOBLF

Nowadays, the practical tasks of UAVs are becoming more and more complicated and diversified. In the practical flight process, the large-scale changes of the flight environment, the modeling errors, and the external disturbances may induce the instability of the UAV flight system. Meanwhile, the constraints of the UAV attitudes also have to be guaranteed during the flight process. However, most existing control methods still have limitations in handling the constraints and the multisource disturbances simultaneously. To address this problem, in this paper, we focus on the actual output tracking control for the UAV systems with full-state constraints and multisource disturbances. Firstly, a high-order tan-type barrier Lyapunov function (HOBLF) has been constructed for the UAV to maintain the full-state constraints. Secondly, by combining the adaptive backstepping technique and the fuzzy logic systems, the modeling errors and the unknown nonlinearities of the UAV attitude control system can be handled. Moreover, by properly constructing several adaptive laws, the time-varying disturbances existing in the UAV attitude control system can be suppressed. Finally, the full-state-constrained antidisturbance controller is formed, ensuring that the tracking error approaches arbitrarily to small neighborhood and does not violate the given constraints. The simulation results illustrate the feasibility and the advantages of the proposed method.

Improving Energy Efficiency in UAV Attitude Control using Deep Reinforcement Learning

Improving Energy Efficiency in UAV Attitude Control using Deep Reinforcement Learning

A Unified SO(3) Approach to the Attitude Control Design for Quadrotors

This paper proposes a two-phased problem reformulation approach to solve the quadrotor UAV attitude control problem. Given an initial attitude, the objective is to design a control to drive it toward a desired value. The first phase begins with the definition of three types of errors pertaining to attitude discrepancy, followed by the utilization of virtual control with which the dynamics of quadrotor’s attitude and angular velocity are then unified under the framework of SO(3), rendering the original control problem much more transparent to tackle. Through cancellation of certain unwanted terms emerging from the unification process, entry-wise treatment of the remaining matrix dynamics is then employed in the second phase to transform it further into the stabilization problem of a 3-dimensional linear time-invariant system that is fairly easy to solve since to which standard feedback designs such as linear quadratic regulator and pole placement are readily applicable. In addition to the classical designs, a specialized and structurally simple controller is provided to reduce the number of gains. To account for system parametric uncertainties, external disturbances, and sensor measurement errors and noises, a second controller based on classical $H_{\infty }$ theory is presented to enhance the robustness of the design. A numerical example is given, and computer simulations are conducted to generate error and control trajectories to assess and compare the effectiveness of the presented designs.

Reinforcement Learning for UAV Attitude Control

Autopilot systems are typically composed of an “inner loop” providing stability and control, whereas an “outer loop” is responsible for mission-level objectives, such as way-point navigation. Autopilot systems for unmanned aerial vehicles are predominately implemented using Proportional-Integral-Derivative (PID) control systems, which have demonstrated exceptional performance in stable environments. However, more sophisticated control is required to operate in unpredictable and harsh environments. Intelligent flight control systems is an active area of research addressing limitations of PID control most recently through the use of reinforcement learning (RL), which has had success in other applications, such as robotics. Yet previous work has focused primarily on using RL at the mission-level controller. In this work, we investigate the performance and accuracy of the inner control loop providing attitude control when using intelligent flight control systems trained with state-of-the-art RL algorithms—Deep Deterministic Policy Gradient, Trust Region Policy Optimization, and Proximal Policy Optimization. To investigate these unknowns, we first developed an open source high-fidelity simulation environment to train a flight controller attitude control of a quadrotor through RL. We then used our environment to compare their performance to that of a PID controller to identify if using RL is appropriate in high-precision, time-critical flight control.

Neural Network Control System of UAV Altitude Dynamics and Its Comparison with the PID Control System

This article proposes a comparative method to assess the performance of artificial neural network’s direct inverse control (DIC-ANN) with the PID control system. The comparison served as an analysis tool to assess the advantages of DIC-ANN over conventional control method for a UAV attitude controller. The development of ANN method for UAV control purposes arises due to the limitations of the conventional control method, which is the mathematical based model, involving complex expression, and most of them are difficult to be solved directly into analytic solution. Although the linearization simplified the solving process for such mathematical based model, omitting the nonlinear and the coupling terms is unsuitable for the dynamics of the multirotor vehicle. Thus, the DIC-ANN perform learning mechanism to overcome the limitation of PID tuning. Therefore, the proposed comparative method is developed to obtain conclusive results of DIC-ANN advantages over the linear method in UAV attitude control. Better achievement in the altitude dynamics was attained by the DIC-ANN compared to PID control method.

Cascade attitude control of a quadcopter in presence of motor asymmetry

The quadcopter is one of the most used unmanned aerial vehicles in both military and commercial in- and outdoor applications. In this study, the problem of UAV attitude control is investigated when there are discrepancies in the characteristics of the 4 actuators, i.e., electrical motors and propellers. To tackle this problem in a simple way, a cascade control strategy is proposed with a PD controller in the inner loop to achieve stabilization, and PI controller in the outer loop to ensure disturbance rejection. This way, the external disturbance created by the actuator asymmetry is compensated by the PI loop. The robustness of the control strategy is tested in simulation as well as in real-life tests on a 1-DOF test bed.

External Perturbation Estimation based on Super-Twisting Algorithm for Attitude Control of UAVs

In this paper, the problem of UAVs attitude control with external perturbation is addressed. The algorithm is constructed in two steps, in the first step the external perturbation is estimated using finite time convergent observer (Super-Twisting Algorithm (STA)) while in the second step this estimation is used by the controller in order to compensate it. The closed loop stability properties of this nonlinear controller/observer are analyzed using separation principle and the system is found to be globally stable. The results of simulations and actual experiments on a real quadrotor show the effectiveness of the proposed observer/controller to improve the performances of the system despite of disturbances.