Organizations: The authors are with the School of Aeronautics and Astronautics, Purdue University, West Lafayette, IN 47906. · The authors are with the Toyota Research Institute of North America, Ann Arbor, MI 48106.
Designing real-time control for nonlinear systems to track periodic orbits can be challenging, especially for complex and coupled dynamical systems (e.g., tethered kites), where the true model parameters are hard to obtain. To address this, we propose a physics-informed Koopman learning approach that incorporates periodic-orbit stability directly into the learned linear predictor. The approach combines the underlying periodic structure of the dynamics along with periodic state-input trajectory data to learn physically consistent linear predictors (in lifted/observable space). We leverage Floquet theory, which studies the stability of periodic orbits in nonlinear systems, to obtain the periodic stability information of a given orbit via a monodromy matrix (i.e., the state transition matrix over an entire period). Subsequently, we impose open-loop monodromy constraints to learn Floquet-parameterized linear predictors using a least-squares Extended Dynamic Mode Decomposition (EDMD) formulation, thereby constraining the search space and aiding in Koopman learning. We then utilize these linear predictors to design a model predictive control (MPC) law for tracking periodic references. Finally, we evaluate the designed framework on a cartpole system under periodic forcing and compare it against a purely data-driven baseline.
Figures & tables
Fig. 1 : Illustration of Floquet-parameterized EDMD ( right ) and purely EDMD ( left ). The main idea is to constrain the search space using Floquet multipliers along a periodic orbit.
Fig. 2 : A cartpole system under a forcing periodic input.
Fig. 3 : Comparison between Vanilla and the proposed Floquet-parameterized EDMDc. Left: Difference between the dominant Floquet multiplier of the learned system and the true system, Center: Worst-of-all difference between the true and estimated Floquet multipliers, and Right: One-step root-mean-squared prediction error between linear predictor and true model.
Method/Snapshots Per Phase
10
20
40
80
160
Vanilla EDMDc [ 5 ]
1507.7
77.6
74.1
73.4
72.9
Floquet EDMDc ( Proposed )
73.7
73.3
73.1
71.9
71.6
TABLE I : Median Root Mean Square Error (RMSE) of the cart position (in mm) for 100 Monte Carlo trials with perturbed initial conditions, comparing both methods trained using various training data sizes per phase bin.