Control system for a self-balancing robot

Abstract

The control of a self-balancing robot has been studied for many years and, despite several achievements, there are still open issues. The aim of this project is to study the efficiency of different control algorithms, as Proportional, Integral and Derivative (PID), pole placement, adaptive control, among others, in a home-made robot called Bimbo. It was also tested an algorithm, applied to the position. The robot was constructed with the modules for movement and position control. It was applied a Kalman filter to get the Roll angle from the Inertial Measurement Unit (IMU). Furthermore, the mechanism to read encoders and to control the two motors was implemented. It was built a mechanism to control system variables through Bluetooth communication, which allows to continuously monitor any robot variable, allowing to test the system and the control variables in real-time. The selected solution implements a PID, continuously changing the reference by a “position algorithm”. A human control interface was created to command Bimbo navigation direction.