NeuroBEM: Hybrid Aerodynamic Quadrotor Model

Bauersfeld, Leonard; Kaufmann, Elia; Foehn, Philipp; Sun, Sihao; Scaramuzza, Davide

doi:10.15607/rss.2021.xvii.042

Cited by 59 publications

(41 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To validate the success rate of navigating through given waypoints in a cluttered environment, we utilize a high-fidelity simulation which is based on Bade-Element-Momentum (BEM) theory [45]. Compared to the simple simulation, the BEM simulation can accurately model lift and drag produced by each rotor from the current ego-motion of the platform and the individual rotor speeds.…”

Section: B Success Rate Of Minimum-time Flight In High Fidelity Simul...mentioning

confidence: 99%

Minimum-Time Quadrotor Waypoint Flight in Cluttered Environments

Pěnička

Scaramuzza

2022

IEEE Robot. Autom. Lett.

Self Cite

View full text Add to dashboard Cite

We tackle the problem of planning a minimum-time trajectory for a quadrotor over a sequence of specified waypoints in the presence of obstacles while exploiting the full quadrotor dynamics. This problem is crucial for autonomous search and rescue and drone racing scenarios but was, so far, unaddressed by the robotics community in its entirety due to the challenges of minimizing time in the presence of the non-convex constraints posed by collision avoidance. Early works relied on simplified dynamics or polynomial trajectory representations that did not exploit the full actuator potential of a quadrotor and, thus, did not aim at minimizing time. We address this challenging problem by using a hierarchical, sampling-based method with an incrementally more complex quadrotor model. Our method first finds paths in different topologies to guide subsequent trajectory search for a kinodynamic point-mass model. Then, it uses an asymptotically-optimal, kinodynamic sampling-based method based on a full quadrotor model on top of the point-mass solution to find a feasible trajectory with a time-optimal objective. The proposed method is shown to outperform all related baselines in cluttered environments and is further validated in real-world flights at over 60km/h in one of the world's largest motion capture systems. We release the code open source.

show abstract

Section: B Success Rate Of Minimum-time Flight In High Fidelity Simul...mentioning

confidence: 99%

Minimum-Time Quadrotor Waypoint Flight in Cluttered Environments

Pěnička

Scaramuzza

2022

IEEE Robot. Autom. Lett.

Self Cite

View full text Add to dashboard Cite

show abstract

“…Due to the high sample complexity of learning-based policies, they are often trained in simulation, which then requires transferring the policy from simulation to the real world. This transfer between domains is known to be hard and is typically approached by increasing the simulation fidelity [10], [11], by randomization of dynamics [6], [12] or rendering properties [13], [14] at training time, or by abstraction of the policy inputs [2], [4]. Apart from simulation enhancements and input abstractions, also the choice of action space of the learned policy itself can facilitate transfer.…”

Section: Supplementary Materialsmentioning

confidence: 99%

“…Experiments are performed on the test trajectories in two settings: (i) in the Nominal setting, the test environment perfectly matches the training environment; (ii) in the Model Mismatch setting, the environment at test time is different from the training environment. Specifically, we use in setting (ii) a quadrotor simulation that was identified from real flight data and uses blade-element momentum theory to accurately model the aerodynamic forces acting on the platform [11]. We also apply a control delay of 20 ms to simulate wireless communication latency.…”

Section: A Simulation Experimentsmentioning

confidence: 99%

A Benchmark Comparison of Learned Control Policies for Agile Quadrotor Flight

Kaufmann¹,

Bauersfeld²,

Scaramuzza³

2022

Preprint

Self Cite

View full text Add to dashboard Cite

“…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 physics-based 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.…”

Section: Introduction a Motivationmentioning

confidence: 99%

“…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].…”

Section: Introduction a Motivationmentioning

confidence: 99%

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

Hanover,

Foehn,

Sun

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Agile quadrotor flight in challenging environments has the potential to revolutionize shipping, transportation, and search and rescue applications. Nonlinear model predictive control (NMPC) has recently shown promising results for agile quadrotor control, but relies on highly accurate models for maximum performance. Hence, model uncertainties in the form of unmodeled complex aerodynamic effects, varying payloads and parameter mismatch will degrade overall system performance. In this paper, we propose L1-NMPC, a novel hybrid adaptive NMPC to learn model uncertainties online and immediately compensate for them, drastically improving performance over the non-adaptive baseline with minimal computational overhead. Our proposed architecture generalizes to many different environments from which we evaluate wind, unknown payloads, and highly agile flight conditions. The proposed method demonstrates immense flexibility and robustness, with more than 90% tracking error reduction over non-adaptive NMPC under large unknown disturbances and without any gain tuning. In addition, the same controller with identical gains can accurately fly highly agile racing trajectories exhibiting top speeds of 70 km/h, offering tracking performance improvements of around 50% relative to the non-adaptive NMPC baseline. We will release our code fully open-sourced upon acceptance.

show abstract

NeuroBEM: Hybrid Aerodynamic Quadrotor Model

Cited by 59 publications

References 28 publications

Minimum-Time Quadrotor Waypoint Flight in Cluttered Environments

Minimum-Time Quadrotor Waypoint Flight in Cluttered Environments

A Benchmark Comparison of Learned Control Policies for Agile Quadrotor Flight

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

Contact Info

Product

Resources

About