Dynamics of a Differential Drive Robot

Autonomous mobile robots rely on dynamic (continuously adapting) system models to locate themselves. Using dual encoders that can measure distance traveled and an IMU (inertial measurement unit) that measures the robot rotation, we can model the robot’s 2D pose (x, y, heading) with ordinary differential equations (ODEs). Furthermore, variations in sensor accuracy, wheel radius, and other parameters can also be studied. Through simulations and real-world testing, we are able to see the strengths and weaknesses of the ODE model when noise and inaccuracy is introduced into the system.