Machine learning aplicado à simulação e controle de um misturador

Abstract

Considering the historical evolution of automation and the increasing complexity of industrial processes, the strategic importance of PID controllers, widely used in regulating critical variables across various industries, is emphasized. The contemporary need for continuous monitoring in chemical facilities and the pursuit of greater operational efficiency drive the investigation of advanced techniques such as machine learning, which has the ability to extract meaningful insights from large volumes of data. This research project aims to develop everything from a detailed simulation of the physical mixer system to the creation of machine learning models specifically for tuning PID controller constants. In summary, the project seeks not only to optimize mixing processes but also to expand knowledge at the intersection of automation, process control, and machine learning, with the potential for significant impact on industrial automation efficiency.