A computer-based fuzzy temperature controller for environmental chambers

Abstract

This paper describes a computer-based temperature control system for an environmental chamber. The identification of the thermal process was made from the analysis of the collected output data in response to random generated input signals. Using this model the optimal linear and nonlinear control strategies were established in order to achieve the desired set-point and to minimize the number of actuations. The actuation over the heating and cooling systems were performed with conventional PID and fuzzy controllers and their responses were compared. This work enhances fuzzy controllers application with self-learning capability to achieve the prescribed control objectives in a near-optimal manner. By applying back-propagation through time it was possible to force the plant block to generate the desired trajectory.