Implementation of Self-balancing Scooter

Abstract

The project is aimed to implement a balancing scooter that can be ride by a person. And the microcontroller unit used in the project is an ARM base core, STM32F4 Discovery board. The balancing algorithm of this scooter is different from the normal balancing robot, because the scooter need to balance itself and the person on it, while a balancing robot only need to balance itself. As the scooter need to consider the person on it, the balancing response of the scooter is different. The transient response of this scooter cannot be too quick, if the scooter reacts too fast, the person may loss its balance. So it is important to make the scooter react reasonably, and then to keep a constant speed at certain angle and move around. By using gyroscope on the board and Hall Effect sensor in the motor, the MCU will capture the data and process the balancing algorithm. In my design, I also implemented a center weight capture module to help to turn the board, and with all the above contribution, my scooter can now be balanced and controlled by a person. Along the whole development process, the hardest part is to tune the balancing gain, which mainly determine the result of riding experience and also the balancing performance. And finally it is reasonably smooth when having the ride.