A Novel Hybrid Intelligent Classifier to Obtain the Controller Tuning Parameters for Temperature Control

Abstract

This study presents a novel hybrid classifier method to obtain the best parameters of a PID controller for desired specifications. The study presents a hybrid system based on the organization of existing rules and classifier models that select the optimal expressions to improve specifications. The model achieved chooses the best controller parameters among different closed loop tuning methods. The classifiers are based on ANN and SVM. The proposal was tested on the temperature control of a laboratory stove.