Data-Driven MPC for Quadrotors

Abstract

Aerodynamic forces render accurate high-speed trajectory tracking with quadrotors extremely challenging. These complex aerodynamic effects become a significant disturbance at high speeds, introducing large positional tracking errors, and are extremely difficult to model. To fly at high speeds, feedback control must be able to account for these aerodynamic effects in real-time. This necessitates a modeling procedure that is both accurate and efficient to evaluate. Therefore, we present an approach to model aerodynamic effects using Gaussian Processes, which we incorporate into a Model Predictive Controller to achieve efficient and precise real-time feedback control, leading to up to 70% reduction in trajectory tracking error at high speeds. We verify our method by extensive comparison to a state-of-theart linear drag model in synthetic and real-world experiments at speeds of up to 14m/s and accelerations beyond 4g. SUPPLEMENTARY MATERIALVideo: https://youtu.be/FHvDghUUQtc Code: https://github.com/uzh-rpg/data driven mpc I. INTRODUCTIONAccurate trajectory tracking with quadrotors in high-speed and high-acceleration regimes is still a challenging research problem. While autonomous quadrotors have seen a significant gain in popularity and have been applied in a variety of industries ranging from agriculture to transport, security, infrastructure, entertainment, and search and rescue, they still do not exploit their full maneuverability. The ability to precisely control drones during fast and highly agile maneuvers would allow to not only fly fast in known-free environments, but also close to obstacles, humans, or through openings, where already small deviations from the reference have catastrophic consequences.Operating a quadrotor at high speeds and controlling it through agile, high-acceleration maneuvers requires to account for complex aerodynamic effects acting on the platform. These effects are difficult to model, since they consist of a combination of propeller lift and drag dependent on the induced airstream velocity, fuselage drag, and complex or even turbulent effects due to the interaction between the propellers, the downwash of other propellers, and the fuselage.

Performance, Precision, and Payloads: Adaptive Nonlinear MPC for Quadrotors

“…These uncertainties can significantly degrade the performance and reliability of the system and potentially lead to loss of control if not compensated for. High-fidelity physicsbased models can improve control performance, but are often prohibitively expensive to procure and require extensive levels of domain expertise [8][9][10][11]. With the advancements of datadriven methods such as those described in [8][9][10], the costs of obtaining accurate models has been dramatically reduced.…”

“…High-fidelity physicsbased models can improve control performance, but are often prohibitively expensive to procure and require extensive levels of domain expertise [8][9][10][11]. With the advancements of datadriven methods such as those described in [8][9][10], the costs of obtaining accurate models has been dramatically reduced. Models learned from data, however, have a tendency to overfit and can be intractable to update online [11].…”

Systematic Online Tuning of Multirotor UAVs for Accurate Trajectory Tracking Under Wind Disturbances and In-Flight Dynamics Changes

“…The results showed enhanced performance in Lemniscate and circular trajectories reaching velocities up to 5 m/s and the authors manually compensated for time delays in the system. A more recent work from the same group could generalize over all trajectory shapes [13]. The authors used a Model Predictive Controller (MPC) for multirotors with the aerodynamic effects modeled using Gaussian processes (GPs).…”

“…In a similar way to the other feedforward tuning methodologies, a regression algorithm learned a drag model based on flight data. All these methods [13], [15], [17], [18] require extensive data collection and offline optimization to build up a drag or disturbance model, which limits the suitability for real-time adaptation. Also the optimized model can be biased towards the training data, and might under perform for unseen scenarios.…”

“…The work of [12] utilized this linearization approach of aerodynamic effects to develop an in-flight tuner of controller parameters. It is more challenging to achieve the desired trajectory tracking performance at high velocities and accelerations because of the increased aerodynamic effects that are difficult to model and compensate for [13]. Thus, aerodynamics accurate modeling of a multirotor is a research challenge, and proper knowledge of such dynamics is essential to guarantee high performance.…”