Melhoria na eficiência energética de um refrigerador termoelétrico: controle preditivo baseado em modelo aplicado à regulação da temperatura do reservatório de água gelada

Abstract

The rational use of resources is a recurring theme in sustainable development discussions and has stimulated the development of more efficient electrical power consumption appliances. Regulators and Certifiers agencies set standards for the functionality requirements and for the electric power consumption classification of devices in order to formalize the relationship between performance and maintenance of these resources indefinitely. Thus a model based predictive controller was applied to a thermoelectric cooler to attain the international standard ANSI18 specifications for water cooler dispensers and to meet ENERGY STAR$ iny^{ extregistered}$ energy efficiency requirements. A phenomenological linear model was obtained and the respective parameters were identified by using least squares algorithm; experiments were performed to identify the parameters of the discrete transfer functions; a model criterion selection based on simulation capabilities and residue characteristics was developed; a model based predictive controller was designed and applied using one of the models identified; the controller performance was analyzed attending to the ANSI18 standard and according to energy performance certificates; a comparative study between the bistable control and the proposed strategy was made for performance analysis. Simulation and experimental tests have shown compromise between the simulation capability and residue characteristic criterion. The closed loop system behavior allowed energy saving in standby temperature regulating compared to thermostat control mode. The new control strategy achieved ENERGY STAR energy-efficiency criteria certification requirements infeasible for traditional regulation.