We present the first perception-aware model predictive control framework for quadrotors that unifies control and planning with respect to action and perception objectives. Our framework leverages numerical optimization to compute trajectories that satisfy the system dynamics and require control inputs within the limits of the platform. Simultaneously, it optimizes perception objectives for robust and reliable sensing by maximizing the visibility of a point of interest and minimizing its velocity in the image plane. Considering both perception and action objectives for motion planning and control is challenging due to the possible conflicts arising from their respective requirements. For example, for a quadrotor to track a reference trajectory, it needs to rotate to align its thrust with the direction of the desired acceleration. However, the perception objective might require to minimize such rotation to maximize the visibility of a point of interest. A model-based optimization framework, able to consider both perception and action objectives and couple them through the system dynamics, is therefore necessary. Our perception-aware model predictive control framework works in a receding-horizon fashion by iteratively solving a non-linear optimization problem. It is capable of running in real-time, fully onboard our lightweight, small-scale quadrotor using a low-power ARM computer, together with a visual-inertial odometry pipeline. We validate our approach in experiments demonstrating (I) the conflict between perception and action objectives, and (II) improved behavior in extremely challenging lighting conditions. SUPPLEMENTARY MATERIAL Video: https://youtu.be/9vaj829vE18 Code: https://github.com/uzh-rpg/rpg_mpc * These authors contributed equally to this manuscript.
This paper proposes a novel control framework that combines the recently reformulated incremental nonlinear dynamic inversion with (higher-order) sliding mode controllers/observers, for generic multi-input/multi-output nonlinear systems, named incremental sliding mode control. As compared to the widely used approach that designs (higher-order) sliding mode controllers/observers based on nonlinear dynamic inversion, the proposed incremental framework can further reduce the uncertainties whilst requiring less model knowledge. Since the uncertainties are reduced in the incremental framework, theoretical analyses demonstrate that the incremental sliding mode control can passively resist a wider range of perturbations with reduced minimum possible control/observer gains. These merits are validated via numerical simulations for aircraft command tracking problems, in the presence of sudden actuator faults and structural damages.
This paper deals with aircraft trajectory control in the presence of model uncertainties and actuator faults.Existing approaches, such as adaptive backstepping and nonlinear dynamic inversion with online model identification, can be applied. However, since there are a number of unknown aerodynamic derivatives, the tuning of parameter update law gains is time-consuming. Methods with online model identification require excitation and the selection of a threshold. Furthermore, to deal with highly nonlinear aircraft dynamics, the aerodynamic model structure needs to be designed. In this paper, a novel aircraft trajectory controller, which uses the Incremental Nonlinear Dynamic Inversion, is proposed to achieve fault-tolerant trajectory control. The detailed control law design of four loops is presented. The idea is to design the loops with uncertainties using the incremental approach. The tuning of the approach is straightforward and there is no requirement for excitation and selection of a threshold. The performance of the proposed controller is compared with existing approaches using three scenarios. The results show that the proposed trajectory controller can follow the reference even when there are model uncertainties and actuator faults.
As a sensor-based control approach, the Incremental Nonlinear Dynamic Inversion (INDI) method has been successfully applied on various aerospace systems and shown desirable robust performance to aerodynamic model uncertainties. However, its previous derivations based on the so-called time scale separation principle is not mathematically rigorous. There also lack of stability and robustness analysis for INDI. Therefore, this paper reformulated the INDI control law without using the time scale separation principle and generalized it to not necessarily relative-degree-one problems, with consideration of the internal dynamics. Besides, the stability of the closed-loop system in the presence of external disturbances is analyzed using Lyapunov methods and nonlinear system perturbation theory. Moreover, the robustness of the closed-loop system against regular and singular perturbations is analyzed. Finally, the reformulated INDI control law and main conclusions are verified by a rigid aircraft gust load alleviation problem.Regarding its applications on aerospace systems, the INDI method is generally used for the inner loop angular velocity control [9][10][11][12], which leads to a relative-degree-one problem for each control channel. The internal dynamics are then avoided by using cascaded control structure, which is a common practice in rigid aircraft flight control designs [10,11,14]. However, the stability of the cascaded control structure is not easy to prove because of its dependency on the time scale separations between different control loops. Also, this cascaded control structure is not suitable for some problems. e.g. It is neither physically meaningful nor practical to separate the higher-order elastic dynamics into cascaded loops. In view of these reasons, the INDI control will be broadened into not necessarily relative-degree-one problems in this paper with consideration of the internal dynamics.The existing derivations of the INDI control law are based on the so-called time scale separation principle, which claims that when the sampling frequency is high, the controls can change significantly faster than the states [9][10][11][12][13][14][15][16][17]. The nonlinear plants are then simplified into linear incremental dynamic equations by omitting state variation related nonlinear terms and higher-order terms in their Taylor series expansion, based on which the incremental control inputs are designed. This approach is not mathematically rigorous since the plant simplification is made before introducing the INDI control inputs and thus becomes deficient for unstable plants. Moreover, although the state related nonlinear terms and higher-order terms are not used in the INDI controller design, they should be kept in the closed-loop dynamic equations and remain influencing the closed-loop system stability and performance, which is also not the case in the existing derivations. Therefore, in this paper, the INDI control law will be reformulated without using the time scale separation principle, and the influences o...
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.