Skip to content

mayiru1021/Koopman-based-MPC

Repository files navigation

Koopman-based Data-Driven MPC

Paper on the way. This repository contains the experiment results and related functions described in the paper "[Paper Title]" by [Author(s)]. The functions that help implement algorithms are given, inlucing kernel functions and hankel matrix functions.

Table of Contents

Introduction

This project implements the method proposed in "[Paper Title]" to [briefly describe the purpose of the project]. This method aims to presents am MPC scheme that ensures recursive feasibility and practical stability for nonlinear systems based on the Koopman operator, extending the aforementioned fundamental lemma to encompass nonlinear systems.. For more details, please refer to our paper.

Requirements

The code is written in MATLAB and requires the following dependencies:

Functions

rbf.m

contains the common kernekl functions including: thinplat BF kernel, gauss kernel, inverse quadratic kernel, inverse multiquadratic kernel and polynomial kernel. The function reference paper "Robust tube-based model predictive control with Koopman operators" by "Xinglong Zhang, Wei Pan, Riccardo Scattolini, Shuyou Yu and Xin Xu".

Hankel.m

can contruct Hankel matrix satisties Wiiliams' Fundamental Lemma.

Results

VandelPol_control_data.mat

control process related data of Van del Pol oscillator.

VandelPol_predict_data.mat

prediction process related data of Van del Pol oscillator only using the multi-step predictor.

non_affine_control data.mat

control process related data of non-affine system.

Uinit_vandelpol.mat

random input sequence for Van del Pol oscillator.

Uinit_nonaffine.mat

random input sequence for non-affine system.

Usage

run main_plot.mat to visualize the results functions: rbf.m and Hankel.m can be used in projects about kernel methods and Hankel matrix.

Citation

@article{yourpaper, title={Paper Title}, author={Author(s)}, journal={Journal Name}, year={Year}, volume={Volume}, pages={Pages} }

License

This project is licensed under the MIT License.

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages