Furuta pendulum
·To refresh my knowledge about dynamical systems and to try out different control approaches in practice I decided to build Furuta Pendulum. I started the project with simulation, in this part I created LQR, MPC and RL controllers and tested them on simulation (Gazebo), about this part of the project you can find out more in my blog post Furuta Pendulum Simulation.
Currently I’m transitioning all the code to the pendulum that I built. This is a work in progress, I plan to update this series so stay tuned. Meanwhile, here you can see the first successful run (LQR + Swing-up) and a more recent update with the Deep Reinforcement Learning controller: