Neural-Network-Based and Robust Model-Based Predictive Control of a Tubular Heat Exchanger

Abstract

The paper is devoted to advanced control of a tubular heat exchanger with focus to energy savings. The controlled tubular heat exchanger (HE) was used for petroleum pre-heating by hot water. The controlled output was the measured temperature of the petroleum in the output stream and the control input was the volumetric flow rate of hot water. Two advanced control strategies were investigated in the set-point tracking, the neural-network-based predictive control and the robust model-based predictive control with integral action and with soft constraints on control inputs. The advanced control of the heat exchanger was implemented in the MATLAB/Simulink simulation environment. Simulation results obtained using advanced controllers were compared with the results ensured by a conventional PID controller and they confirmed significant improvement of the control performance. Moreover, advanced controllers reduced energy consumption measured by the total consumption of hot fluid used for heating.