Paper Conference

2020 Building Performance Analysis Conference and SimBuild co-organized by ASHRAE and IBPSA-USA


Decentralized Approach to Multi-Zone Grey-Box Modeling for Model-Based Predictive Control

Jaewan Joe, Borui Cui, Piljae Im, Jin Dong, Kuruganti Teja
Oak Ridge National Laboratory, Oak Ridge, United States of America

Abstract: This study aims to improve the easiness of utilizing the grey-box model (i.e., Resistance-Capacitance circuit) for Model-based Predictive Control (MPC). The primary barrier of implementing the MPC is estimating the control-oriented building model that needs to be computationally inexpensive and quick but reasonably precise in predicting building load and indoor conditions. The estimating of the model parameters becomes more complicated when the building scale is larger; e.g., multi-zone building. In this study, a decentralized approach is introduced; each zone is split and individually estimated with measured boundary temperature from adjacent zones integrated into one single system model. The proposed decentralized method is demonstrated with experimental data from a full-scale multi-zone test cell compared with the centraliozed reference case.
Pages: 545 - 551