Real-Time Adaptive Predictive Control for Hybrid Thermoelectric Refrigerator System

142803-7272-IJMME-IJENS © June 2014 IJENS I J E N S  Abstract– This paper reports on control studies for Hybrid Thermoelectric Refrigerator (H-TER) system. The system is highly non-linear and exhibits varying model parameters and dead-time, hence the objective of the study is to investigate control strategies that are not based on detailed advance plant knowledge but allow for continuous adaptation of the controller to changing system dynamics. Using the generalized predictive controller (GPC), it is observed that the H-TER system is always stable even in the presence of randomly varying or step disturbance as well as dynamics model. The use of tuning parameters serve to improve system performance and to limit the controller from producing undesirable and unrealistically large, fast varying control signals. An interesting outcome from these studies is that controller design using GPC provides a more systematic approach is choosing the design parameters to achieve the desire performance specifications. However, it is also noted that the effects of non-linearity and varying time delay has not been investigated for this system. Under such situations, the design must involve the choice of prediction and control horizon, assumed to be fixed in this study. Then, it would mean more computation as the matrices involved would considerably increase in dimension.