This notebook builds upon what has been described in Part I. In Part I, we introduced the linear–quadratic regulator (LQR) framework in Python. We solved the linearized control problem.
In this notebook, we will see that we can do better. The basic idea is to follow the the evolution of “observables” — functions of the state space — instead of the evolution of the state itself using the Koopman operator.
The two main goals of this blog post is to introduce what the linear–quadratic regulator (LQR) framework is and to show how to solve LQR problems using Python. The LQR is concerned with operating a dynamic system (a rocket, a car, the economy, etc.) at minimum cost.
In this blog post you will learn what the LQR framework is how to simulate forward an ordinary differential equation using scipy how to solve for the optimal control using the Python Control Systems Library The Jupyter notebook with the code used to generate this blog post can be found here