Navigation of Self-Balancing Mobile Robot through Sensors

摘要

The aim of this paper is the design, simulation, construction and programming of a self-balancing (inverted pendulum) mobile robot, capable of obstacle avoidance through the MPU6050 sensor, and ultrasonic sensor. The robot will only use a MPU6050 sensor for evaluating and navigating the terrain, a triple axis accelerometer and gyro for calculating angular velocities and accelerations, and Arduino. The design of the robot is performed in CAD-Inventor and Fusion360. The simulation is done mainly on Matlab. The mechanical parts have been printed at the 3D printer that we developed at our university. The modeling of the robot is performed using the inverted pendulum as a reference and a space state model of the robot is generated. This comes as a natural requirement in order to use the Linear Quadratic Regulator and Kalman filter, as the main process controller of this system. The required states are obtained by processing the MPU6050 sensor. This information is used to calculate the feedback controller gains, in order to have an overall better system response. The application of the algorithm after some of the examinations made has shown precision in detecting objects and finding the path for the fastest time.