Two-Wheeled Self Balancing Robot

Abstract

Attached

This paper presents a method to design and control a two-wheeled self-balancing robot by focusing on hardware description, Complimentary filter algorithm, system modelling and Proportional-Integral-Derivative (PID) back stepping controller design. In the system, signals from the digital Gyro sensor are filtered by a Complimentary filter before being fed to the Proportional-Integral-Derivative back stepping controller. The objectives of the proposed controller is to stabilize the robot while trying to keep the motion of robot to any direction. By experimenting, the values o f Proportional-Integral-Derivative parameters such as, Proportional Gain Constant, Derivative Gain Constant and Integral Gain Constant have been obtained and applied for the Arduino board. The special software was complied to convert the digital data from the accelerometer to an acceleration magnitude vector. The magnitude is then compared to a predetermined mathematical function to infer the angle o f tilt of the platform. The angle of tilt is then converted to angle of rotatior. for the gear motors to act on. Complimentary filter is used to filter the gyro data in order to reduce noise, drift, and horizontal acceleration dependency, for the fast estimation of angle. It was noticed that this approach promises much less lag than the low-pass filter alone, and not very processor-intensive. Tuning for Proportional-Integral-Derivath e c