Python library for analysis of time series data including dimensionality reduction, clustering, and Markov model estimation

A complete, hardware-ready Python package for Koopman-based Linear Model Predictive Control (LMPC), delivering real-time trajectory tracking for quadrotors using analytical Koopman lifting (no training data required)

This document explains the implementation of the Koopman Operator in conjunction with Model Predictive Control (MPC) to control a nonlinear system.

Extended Dynamic Mode Decomposition for system identification from time series data (with dictionary learning, control and streaming options). Diffusion Maps to extract geometric description from data.

A repository for an online adaptive Koopman algorithm described for the paper titled "Adaptive Koopman Architectures for Control of Complex Nonlinear Systems".

This repository contains all the work developed in the context of the Master Thesis dissertation entitled Model Predictive Control for Wake Steering: a Koopman Dynamic Mode Decomposition Approach. The repository includes all developed documentation (dissertation, extended abstract, poster and presentation) source code (MATLAB script and function…

My Master Thesis in the area of Data-Driven Control Engineering

This code can be used to reproduce the results in our paper ``Extended Kalman filter---Koopman operator for tractable stochastic optimal control'.

Scaling Law of Neural Koopman Operators

Official repository for CoRL 2025 Paper: 'KoopMotion: Learning Almost Divergence Free Koopman Flow Fields for Motion Planning' by Li et al. from GRASP Lab, UPenn

Official code for the IEEE SPL paper "Stabilizing RED using the Koopman Operator (SKOOP-RED)." Includes implementations, demos, and scripts to reproduce results and plots.

koopman operator examples

Koopman-Assisted Reinforcement Learning

A framework for data-driven modeling and analysis of granular materials in the strongly nonlinear regime using the modern Koopman theory

A Graph Dynamical Neural Network Approach for Decoding Dynamical States in Ferroelectrics.

Stability-guaranteed network for learning Complex dynamical systems

SPIKE: Sparse Koopman Regularization for Physics-Informed Neural Networks

Koopman-Assisted Reinforcement Learning

Physics-informed hybrid quantum-classical autoencoder that learns the Koopman operator for chaotic fluid dynamics (Kuramoto-Sivashinsky). Fuses symplectic geometry with CV photonic quantum circuits (PennyLane) to guarantee zero energy drift over 100K+ recursive time steps.

Adaptive deep Koopman predictive control for mitigating traffic oscillations in mixed CAV/HDV traffic, with pretrained checkpoint and reproducible simulation code.