Linear Quadratic Regulator (LQR) Optimal Control of a Differential Drive Robot

The LQR control algorithm was used to generate optimal control input to send a Husarion robot to a goal pose by minimizing the state and input cost function. The desired goal was set to (x=1.0, y=1.0, theta=1.57).

Setup

• Get the robot

mkdir -p ~/robot_ws/src
cd ~/robot_ws/src
git clone -b foxy https://github.com/husarion/rosbot_description.git

• Get the LQR

cd ~/robot_ws/src
git clone https://github.com/siksal/control.git

• Build the packages

cd ~/robot_ws
colcon build
source install/setup.bash

Simulation

• Run the LQR for Rosbot

## Launch robot
ros2 launch rosbot_description rosbot_sim.launch.py

## Run LQR in another terminal/tab (remember to source the new terminal/tab)
ros2 run control lqr

Result:

lqr.mp4