SUPERVISED TUNING OF A CHILLER WITH A COLD PHASE-CHANGE MATERIAL USING AN RSM-SMITH PREDICTOR

Abstract

The proposed work is focused on the tuning of the latent-heat storage of a PCM with MWCNT encapsulation for a chiller system in the milk-pasteurization process using an RSM-based predictive Smith controller. The PCM was synthesized using sodium polyacrylate and MWCNT particles, encapsulated in spherical balls stored in the IBT (ice bank tank of the chiller unit). Experimental work is conducted on the heat-transfer characteristics of the chiller unit with servo-operated flow-control valves based on the central composite design of experiments. The system is cascaded using a PID controller with a Smith predictor to stabilize the latent temperature of milk cooling in the chiller system. The system is considered to be a first-order transformation with the rise and settling time. The plant model is obtained, using the response-surface method based on the transfer function to minimize the error derivatives of the chiller system. Results of the proposed method show a robust performance in keeping the PCM temperatures and bacterial-growth value stabilized, allowing a less drastic control compared to the cascade-model predictive control.