Study On Model Predictive Control Of Plate-finHeat Exchangers

Abstract

Heat exchangers are important pieces of equipment to transfer heat between process fluid media. In some processes the working fluid temperatures must be exactly controlled. The common control method such as PID feedback control might not provide high control quality because such a control is based on the deviation between inputs and outputs, and therefore the disturbances in the outputs cannot be eliminated if disturbances in the inputs occur. To overcome the disadvantages of traditional control methods, a model predictive control method is developed. A mathematical model for dynamic behaviour of one-dimensional flow multistream plate-fin heat exchangers is proposed to calculate the adjustment parameters. A model predictive control method is put forward by introducing a feed-forward controller. Experiments are carried out to compare the model predictive control with the PID feedback control method, which shows that the adjustment time and fluctuations of the controlled outputs of model predictive control are much smaller than those of PID feedback control.