LPV model-based temperature control of thermoelectric device

A linear parameter varying (LPV) model for a class of nonlinear thermoelectric (TE) system is presented, which provides a relatively simple modeling process and high model accuracy. Identified model parameters are applied to design the adaptive Smith PID controller, which simplifies the computation of the control process owing to the low order of LPV models and the Smith structure. Further more, a multiple reference models (MRM) structure is proposed to design the controller for a class of nonlinear systems with the LPV models like the TE system, introducing a reference model observer (RMO) to generate an appropriate reference model (RM) for the different set-points in order to solve the global tracking problems of the nonlinear systems. The simulation and experimental results show that the proposed model is able to demonstrate the nonlinearity of the TE systems in a wide input domain. The proposed adaptive PID method can also deal with the nonlinear dynamic of the TE systems for the set-point tracking over a large temperature range with improved temperature tracking performances, which reduce the overshoot and improve the transient time.