A model predictive control regulation model for radiant air conditioning system based on delay time

Abstract

Radiant air conditioning (RAC) systems are widely used because of the advantage of higher thermal comfort compared to convection systems. However, RAC systems exhibit considerable hysteresis and significant time delays in heat transmission from the terminal to the interior environment. Traditional feedback control methods do not adequately account for the system's delay impact while implementing dynamic control, resulting in a mismatch between supply and demand loads in time series, an inefficient impact of interior temperature control, and energy waste. According to this study, the time-delay characteristics of the RAC system are quantified by the startup-process delay time (SPDT) and operating-phase integrated delay time (IDT). The system startup time is set by the SPDT. Based on the IDT, the prediction interval of the model predictive control (MPC) and the step time of rolling optimization are determined. On this basis, a short-term forecasting model of air conditioning operating load and the regulation model of the system using MPC are established and verified by comparing with the conventional rule-based control of constant water temperature (CRBC) and proportionally-integral-differential (PID) control. The results for the RAC system with a large time delay indicate that the stability of indoor temperature under MPC controlling is 18.8% higher than the PID control and 63.5% higher than the CRBC method. Also, the indoor temperature controlled by the MPC is closest to the set value and shows the minimum fluctuation. Furthermore, the overall energy consumption of the system is estimated for reducing by 14.89% in comparison with the CRBC and by 8.51% in comparison with the PID.