Control strategies for two-wheeled self-balancing robotic systems based on Kalman filter and LQR

Abstract

This paper purpose a summary of the development process of our two-wheeled self-balancing robot (TWSBR). We explore the fundamentals of locomotion, emphasizing the robot’s inspiration from the principles of an inverted pendulum. Dynamic modeling enables us to understand the robot’s behavior and stability, leading to the implementation of robust linear controllers for balance maintenancewe explore the Linear Quadratic Regulator (LQR), a state-of-the-art linear control strategy highly acclaimed for its effectiveness in managing linear systems. The LQR provides a robust and optimal control approach that harnesses the principles of state-space representation and optimal control theory to achieve the desired behavior of a system. Our exploration focuses on unraveling the underlying principles of the LQR and its practical application in the context of a two-wheel self-balancing robot (TWSBR). Furthermore, we showcase the practical implementation of both the LQR and PID control strategies within the control architecture of the TWSBR, discussing the calibration process and the challenges encountered along the way, including the impact of motor encoder issues. Through this comprehensive exploration, we aim to equip readers with valuable insights into developing robust control systems for self-balancing robots, contributing to the advancement of control methodologies and inspiring further innovation in the field of robotics.