We’ve just released Open Control - a python package that implements basic algorithms for analysis and design of optimal feedback controllers.
Classical control methods
Linear quadratic regulator (
Adaptive Dynamic Programming (
ADP) for optimal linear/nonlinear control systems
Off-policy, On-policy learning algorithms for linear/nonlinear systems
Experience replay algorithms for linear/nonlinear systems (TODO)
OpenControl provides algorithms for wide range of system dynamics, including classical control module and reinforcement learning control systems with ADP module. For learning purpose, we recommend starting with our example code to define a system and run the simulation.
You can check out the latest version of our source code by cloning the repository:
git clone https://github.com/VNOpenAI/OpenControl`
Before all, create a new virtual conda environment and install requirements
pip install requirements.txt
You can do minimal installation of OpenControl with
pip install OpenControl
python >>> import OpenControl >>> OpenControl.__version__
Please read the lastest version of our documentation here
For any question, drop an email at firstname.lastname@example.org.