Desarrollo de bio controladores meta heurísticos con realimentación por visión artificial para un sistema RollingBall and Plate

Abstract

The Rolling/Ball and Plate system has two problems when calibrating the PID controller that will govern the system. The first problem is that the system is non-linear and the second one is that there is a high grade of uncertainty when developing the mathematical model so it is possible to apply the conventional tailor-made solutions like Ziegler Nichols, Root-Locus, and others and expect to have an optimal result. In this study, six relevant bio-algorithms were developed in order to calibrate the Proportional, Integral and Derivative (PID) controllers that always search for the optimal PID controller constant values. Once the results were obtained, a statistical test was applied, finally the tuning of PID controllers by meta-heuristic algorithms was tested in the plant 33-240 of the manufacturer National Instruments with different input signals and their behavior was analyzed in order to compare and obtain conclusions about the performance of each bioalgorithm.