Obtaining Continuous Process Models Using Matlab Systems Identification Toolbox

Abstract

The present work explore with Matlab System Identification Toolbox, it allows quite easily to obtain a model of the process to be tested for its subsequent control, based on the registered data of input and output. These experimental data acquired by an oscilloscope or a data acquisition system serve to obtain the model of the process, but must first be processed using system identification algorithms. To obtain continuous models, we have explored the option: Process Models and Transfer Functions Models of the Matlab Systems Identification Toolbox. It has been proven for our test data that the best option was Process Models. In addition, other options were obtained from this Toolbox, which has allowed us to securely obtain the model that best approaches the test plant, being the validation with data different from the identification, input and output charts, input and exit graphics validation, and others.