Evaluación del desempeño de dos algoritmos de control en tiempo real usando Arduino y FPGA. Caso de estudio: péndulo de Furuta

Abstract

This paper is about the design and implementation of two control algorithms for a Furuta pendulum. The implementation of these control algorithms is being done on two embedded platforms based on the microcontroller Arduino and FPGA. In the first development phase of the work, a theoretical review of the control techniques applied to the pendulum (PID and states feedback) is done. In the second phase, the PID controller is designed and implemented. The calibration of its parameters is done through the trial-error heuristic technique. In the third phase, the controller of the states feedback is performed. This phase begins with the mathematical modeling of the system. Then, the model is validated with Matlab’s help. Once the model has been validated and parameterized, it is linearized, and the feedback vector is obtained with the Ackerman formula. Both controllers are implemented in the Arduino and FPGA embedded platforms. Data from the experiments that were done for these controllers are acquired by a National Instruments acquisition card. Matlab is used to interpret these data, where are digitized and the average error rates, average energy and total energy are calculated. After evaluating the results of the experiments, it is found that both control algorithms fulfill the objective of stabilizing the pendulum. When comparing the results for both the PID controller and the status feedback controller, their indices indicate that the best controlling embedded platform is the FPGA. To decide between which algorithm should be used to control the plant, the amount of technical information available must be considered first.