Fuzzy Models for a Two-Wheeled Self-Balancing Robot

Abstract

A self-balancing robot is a robot that is primarily capable to maintain a stable tilt angle. These kinds of robots have the advantages of having small size, footprint, being easy to carry and practical. The system is mainly unstable by nature, which need control strategies to maintain is equilibrium. The main goal of this paper is the development of fuzzy models for a two-wheeled self-balancing robot, previously built and tested experimentally. The fuzzy model input is the control signal applied to the stepper motor, i.e., the number of steps, and the fuzzy model output is the tilt angle, namely the robot angle with respect to the vertical equilibrium point. An incremental online algorithm is applied in order to obtain evolving fuzzy models. The system usually uses a microcontroller, an accelerometer, and a gyroscope module to obtain the tilt angle, and stepper motors play the role of actuators. The developed fuzzy models are validated against experimental input-output data measured from the real robot.