Estudo experimental de robustez de um controlador PID adaptativo por modelo de referência (PID-MRAC) com entradas do tipo caóticas

Abstract

The present work describes the methodologies used to evaluate the robustness of an adaptive controller used to promote pressure management in a water supply system when the system is subjected to chaotic inputs. Pressure is one of the most important quantities in supply systems and it directly impacts other associated variables such as flow for example. The standardization of pressure in the supply network, through control adjustment, can reduce the frequency of pipe ruptures, pressure-induced excessive water consumption and volume lost in leaks. To this end, the control technique used in this work was adaptive control. The analyzed controller, based on the technique, can change its gains from the dynamic changes that occur in the system over time, without necessarily having a previous data collection for the controller modeling. Tests were performed emulating chaotic inputs, such as spurious outliers and gaussian noises, likely to occur in water supply systems, to verify the controller's robustness to these inputs. The results obtained and analyzed proved the robustness of the controller, allowing a window of applicability in real supply systems.