Design and Modeling of a Two-wheeled Differential Drive Robot

There are many existing publications on balancing two-wheeled, differential drive robot (TWDDR) covering dynamic modeling, kinematic modeling, path planning, control architecture design and/or simulations. However, there are few papers that cover all of these in a comprehensive manner that is approachable to beginner robotics researchers. This paper provides step-by-step details of the robotic design process including dynamic modeling, kinematic modeling, linearization, autonomous navigation, path planning, and stability control. A cascaded PID control architecture is presented that is capable of stabilizing the robot in less than 1 s with minimal steady-state error and performing large force and torque disturbance rejection. Additionally, a high-level path planning algorithm based on artificial potential fields is demonstrated.