Skip to content

Thesis: Application of Reinforcement Learning for the Control of Nonlinear Dynamical Systems

Notifications You must be signed in to change notification settings

mpritzkoleit/pygent

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

pygent

This repository arised out of a diploma thesis at Institute of Control Theory, Dresden University of Technology, Germany. The software led to several publications where it is cited. More information and additional material regarding the the publications can be found in the respective branches:

  • ifac: "Reinforcement Learning and Trajectory Planning based on Model Approximation with Neural Networks applied to Transition Problems"
  • gma: "Bestärkendes Lernen mittels Offline-Trajektorienplanung basierend auf iterativ approximierten Modellen"
  • at-beitrag: "Bestärkendes Lernen mittels Offline-Trajektorienplanung basierend auf iterativ approximierten Modellen"

Thesis: Bestärkendes Lernen zur Steuerung und Regelung nichtlinearer dynamischer Systeme (engl. Reinforcement Learning for the Control of Nonlinear Dynamical Systems)

The goal of the thesis was to investigate the state-of-the-art in reinforcement learning for continuous control of nonlinear dynamic systems

As a reference method, iLQR is implemented, a trajectory optimization algorithm referenced in many RL papers, to solve such problems in a model based fashion.

Installation:

clone or download the package

run: python setup.py install

For now please have a look at the 'examples' folder.

About

Thesis: Application of Reinforcement Learning for the Control of Nonlinear Dynamical Systems

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages