Robust error-based active disturbance rejection control of a quadrotor

Abstract

Purpose The control of a quadrotor unmanned aerial vehicle (UAV) is a challenging problem because of its highly nonlinear dynamics, under-actuated nature and strong cross-couplings. To solve this problem, this paper aims to propose a robust control strategy, based on a concept of active disturbance rejection control (ADRC). Design/methodology/approach The altitude/attitude dynamics of a quadrotor is reformulated into the ADRC framework. Three distinct variations of the error-based ADRC algorithms, with different structures of generalized extended state observers (GESO), are derived for the altitude/attitude trajectory-following task. The convergence of the observation part is proved based on the singular perturbation theory. Through a frequency analysis and a quantitative comparison in a simulated environment, each design is shown to have certain advantages and disadvantages in terms of tracking accuracy and robustness. The digital prototypes of the proposed controllers for quadrotor altitude and attitude control channels are designed and validated through real-time hardware-in-the-loop (HIL) co-simulation, with field-programmable gate array (FPGA) hardware. Findings The effects of unavailable reference time-derivatives can be estimated by the ESO and rejected through the outer control loop. The higher order ESOs demonstrate better performances, but with reductions of stability margins. Time-domain simulation analysis reveals the benefits of the proposed control structure related to classical control approach. Real-time FPGA-based HIL co-simulations validated the performances of the considered digital controllers in typical quadrotor flight scenarios. Practical implications The conducted study forms a set of practical guidelines for end-users for selecting specific ADRC design for quadrotor control depending on the given control objective and work conditions. Furthermore, the paper presents detailed procedure for the design, simulation and validation of the embedded FPGA-based quadrotor control unit. Originality/value In light of the currently available literature on error-based ADRC, a comprehensive approach is applied here, which includes the design of error-based ADRC with different GESOs, its frequency-domain and time-domain analyses using different simulation of UAV flight scenarios, as well as its FPGA-based implementation and testing on the real hardware.

Active Disturbance Rejection Control (ADRC) Toolbox for MATLAB/Simulink

“…Remark 7. The presence of control signal u(t) in the total disturbance d(•), visible in (7), does not cause instability of the system as long as condition (5) is satisfied.…”

“…It is, however, recommended to be as precise as possible when setting the value b(•). Not only it can lead to the system instability when set outside the aforementioned acceptable range, but also can cause the total disturbance to be harder to estimate, since the difference between the real input gain parameter b(•) and its assumed value b(•) is incorporated within the total disturbance, represented by (7). Furthermore, when the value b(•) approaches the boundaries of ( 5), it may cause some unwanted oscillations in the system behavior.…”

“…Analyzing the market and the scientific literature, one may find a variety of such software toolboxes supporting the synthesis of control systems, with emphasis on e.g. identification of dynamic systems [1], modeling and solving optimization problems [2], signal differentiators [3], disturbance observers [4], intelligent control based on brain emotional learning [5]), fractional-order control [6], or rapid prototyping [7].…”

Simplifying ADRC design with error-based framework: case study of a DC–DC buck power converter

“…The error-based ADRC has been successfully applied by now to a variety of control problems including mechanical vibration compensation [21,22], robotic manipulator control [23,24], observer design with suppression of sensor noise [25][26][27], position control of a telescope mount [28], UAV control [29,30], as well as power electronics control [31,32].…”

Robust active disturbance rejection control for systems with internal uncertainties: Multirotor UAV application

“…In this study, we focus on the stability and performances of ADRC for multirotor UAVs when parametric uncertainties are not neglected. ADRC has already been successfully applied for the control of this kind of aerial vehicles, mostly on the attitude control loop like in Castillo et al (2016), Yang et al (2018), Yuan et al (2019) and Castillo et al (2019) or attitude and altitude as in Lechekhab et al (2020), but also for both position and attitude control like in Zhang et al (2018) and Xu et al (2020). It was demonstrated that the uncertainties on model parameters (the uncertainties of mass, inertia and actuation efficiency) can be considered as additional disturbances to be estimated by the observer and that they do not affect the stability under some conditions (existence of upper bounds).…”