Towards Machine Learning-based Model Predictive Control for HVAC Control in Multi-Context Buildings at Scale via Ensemble Learning

TL;DR

This work tackles scalable HVAC control by learning a target-building thermodynamics model through a hierarchical ensemble of existing base models. It introduces ReeM, a two-level HRL framework that first selects a subset of base models and then assigns weights to form an ensemble, aided by a temporal encoder and REINFORCE training. Offline experiments over 65 Osaka University rooms and an on-site case study demonstrate substantial improvements in prediction accuracy and reductions in energy use compared with baselines, while supporting online adaptation without retraining. The approach enables reuse of diverse models across buildings but faces challenges in feature alignment and ensuring physical consistency within the ensemble.

Abstract

The building thermodynamics model, which predicts real-time indoor temperature changes under potential HVAC (Heating, Ventilation, and Air Conditioning) control operations, is crucial for optimizing HVAC control in buildings. While pioneering studies have attempted to develop such models for various building environments, these models often require extensive data collection periods and rely heavily on expert knowledge, making the modeling process inefficient and limiting the reusability of the models. This paper explores a model ensemble perspective that utilizes existing developed models as base models to serve a target building environment, thereby providing accurate predictions while reducing the associated efforts. Given that building data streams are non-stationary and the number of base models may increase, we propose a Hierarchical Reinforcement Learning (HRL) approach to dynamically select and weight the base models. Our approach employs a two-tiered decision-making process: the high-level focuses on model selection, while the low-level determines the weights of the selected models. We thoroughly evaluate the proposed approach through offline experiments and an on-site case study, and the experimental results demonstrate the effectiveness of our method.