The controller implementation of a balancing robot using a system model with acceleration as the input

초록

In this paper, we derive a new type of mathematical model for a two-wheeled balancing robot and propose a controller implementation method based on the model. In the newly derived model, the acceleration of the robot body is used as the control input for the system, which provides the advantage of flexible selection of motors used for driving the robot. First, using Lagrangian dynamics and variable elimination, we derive a new type of mathematical model in which the acceleration of the robot body is the input to the balancing robot. Second, we propose a balancing control method based on the newly derived model. The structure of the controller adopts the LQR (Linear Quadratic Regulator) to determine the acceleration required for balancing. The obtained acceleration is converted to the velocity references of two wheels through integration and transformation. The resultant velocity references are followed through PI-type velocity controllers. Third, the implementation of the controller and the experimental results are presented. Two balancing robots are built, one of which adopts DC motors while the other uses step motors for actuation. Control experiments are conducted to verify the effectiveness of the proposed model and control method. The experimental results are illustrated through a phase portrait and the pitch angle of the robot and the control performance is analyzed. The proposed model and control method do not require the electrical coefficients of the adopted motor, thus offering great flexibility in motor selection. © ICROS 2018.

키워드