Research And Design Of The Control System Of Two-wheeled Robots Based On STM32

Since the American inventor Dean Kamen designed the first balancing robot (Segway) in 1995, it has been great concern by the people. It is characterized by a total of two wheel shaft, differential movement, in accordance with principles of dynamic equilibrium of inverted pendulum. Until now, it is limited to transportation applications, and expensive. In addition, the robot is a nonlinear and unstable system. When it is in balance, it needs to continuously adjust the torque wheel. So the system will be in power state, it will lose balance once it's power lower than required. Therefore, how to make the robot in the feature-rich while reducing costs and how to save power energy. It has become a key topic in this research paper.For the purpose of solving the above problems, this paper presents the balance of the robot based on STM32. The features of the robot are that the system power support can be shut off by remote controller when it is not needed. So that, the systems is in sleep mode to save energy. According to the remote control signal, the robot can automatically restore the balance with a mechanical arm when it is needed. In the function expansion, a LCD display and a open-source real-time operating system,μCOS-Ⅱ, is applied to the robot. This will make room for the future function expansion. Abouting cost control; some components are selected to meet the robot's performance at the same time being lower cost. The specific work has done in the following aspects.1) Four steering servo are used, the two have been transformed into servos used to drive robot's two wheels. The two are used to control mechanical arm for the purpose of restoring the balance of the robot. Those Steering gears are all controlled by steering servo controller.2) A encoder is designed to meet the performance with the limited cost.3) Accelerometers, gyroscopes, remote transceiver, LCD display, steering gear, steering gear control and micro-controller are selected which can meet the performance while lower cost. respectively, schematics and PCB are drew for these components. The robot's entire circuit is builted.4) According to the theory of Kalman filter, Kalman filter code is programmed for flitting the interference. By the Kalman filter, the gyroscope and accelerometer's optimal value of the angular velocity can be found.5) In order to achieve the ideal control effect, PID regulator code is programmed according to the principle of PID regulator.6) Real-time operating systemμCOS-Ⅱhas successfully transplanted into the STM32. robot's tasks and interrupt service routine code are programmed.After the completion of above works, it is connected between the PC and robot's SW interface is by ULINK-Ⅱ. Using ARM Real View MDK's trace function, robot body angle is tracked.results show that the robot running smoothly, and can automatically restore the balance after lost balance. It is achieved the design purpose.