Research on a forecasted load-and time delay-based model predictive control (MPC) district energy system model

Abstract

District energy system is an underground network used to cool and heat buildings efficiently, which has the characteristics of large scale, multiple equipment, large lag, nonlinear and time-varying dynamics. The traditional feedback control strategies not only cause mismatch and hysteresis of energy supply and load demand in time series but also lead to high energy consumption. Therefore, a forecasted load-and time delay-based model predictive control district energy system model is proposed in this paper. Taking a district energy system in Tianjin University as a test case, the superiority of the proposed control model is ascertained. A TRNSYS and MATLAB co-simulation platform is developed as the test platform. The results show that the proposed control model can not only improve matching the extent of energy supply and load demand in time series, but can also reduce the energy peak and total energy consumption of the district energy system. Specifically, energy consumption is reduced by 9.56% and 5.30% in summer and 7.75% and 2.61% in winter compared with the control models based on the traditional feedback control and load feedforward fuzzy control strategies, respectively.