English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Nonlinear dynamical effects are crucial to the operation of many agilerobotic systems. Koopman-based model learning methods can capture thesenonlinear dynamical system effects in higher dimensional lifted bilinear modelsthat are amenable to optimal control. However, standard methods that lift thesystem state using a fixed function dictionary before model learning result inhigh dimensional models that are intractable for real time control. This paperpresents a novel method that jointly learns a function dictionary and liftedbilinear model purely from data by incorporating the Koopman model in a neuralnetwork architecture. Nonlinear MPC design utilizing the learned model can beperformed readily. We experimentally realized this method on a multirotor dronefor agile trajectory tracking at low altitudes where the aerodynamic groundeffect influences the system's behavior. Experimental results demonstrate thatthe learning-based controller achieves similar performance as a nonlinear MPCbased on a nominal dynamics model in medium altitude. However, ourlearning-based system can reliably track trajectories in near-ground flightregimes while the nominal controller crashes due to unmodeled dynamical effectsthat are captured by our method.

Quadrotor Trajectory Tracking with Learned Dynamics: Joint Koopman-based Learning of System Models and Function Dictionaries