Identificação e Controle de um Processo de Nível: Aplicação Experimental da Técnica de Augmentação Estocástica de Controlador PID

Abstract

In this article it is shown the experimental results of the application of a design technique that adds stochastic characteristics to the PID controller, enhancing its capacity to handle disturbances such as process noise and measurement ones. The experiments were conducted on a type of plant known as level tank. The model of the plant was obtained experimentally and based on the ARMAX structure via parametric identification using the MATLAB software. Such model structure can incorporate the process noise in the colored case, allowing the application of the stochastic augmentation of the PID controller by means of the generalized minimum variance control method. Simulations and experimental results obtained on the level tank demonstrated that the stochastic PID case has outperformed the conventional digital PID in terms of robustness to noise, showing better performance and less energy consumption.