This paper introduces a methodology for predicting a warehouse’s reduced load while offering flexibility. Physics-based energy simulations are first performed to model flexibility events, which involve adjusting cooling setpoints with controlled temperature increases to reduce the cooling load. The warehouse building encompasses office and storage spaces, and three cooling scenarios are implemented, i.e., exclusive storage area cooling, exclusive office area cooling, and cooling in both spaces, to expand the study’s potential applications. Next, the simulation data are utilized for training machine learning (ML)-based pipelines, predicting five subsequent hourly energy consumption values an hour before the setpoint adjustments, providing time to plan participation in demand response programs or prepare for charging electric vehicles. For each scenario, the performance of an Artificial Neural Network (ANN) and a tree-based ML algorithm are compared. Moreover, an expanding window scheme is utilized, gradually incorporating new data and emulating online learning. The results indicate the superior performance of the tree-based algorithm, with an average error of less than 3.5% across all cases and a maximum hourly error of 7%. The achieved accuracy confirms the method’s reliability even in dynamic scenarios where the integrated load of storage space and offices needs to be predicted.

The nighttime setback approaches commonly consider fixed morning schedules regardless of the rooms’ thermal behavior, resulting in rooms being conditioned before occupants’ arrival. The present work is focused on developing machine learning-based pipelines for estimating the ramp-up duration in different indoor spaces, which is the time the HVAC system needs to bring the space temperature to the desired setpoint. A medical center located in northern Italy, with an HVAC system monitored during the cooling season, is employed as the case study, where the ramp-up procedure was experimentally deployed and investigated under various climatic conditions. Accordingly, machine learning-based pipelines with six different algorithms for estimating the ramp-up of each space are developed, permitting just-in-time start-up of the fan coil units for each day. Next, the most influential features are specified for each room’s best-performing model, aiming for performance improvement and computational cost reduction. The obtained results indicated an average Mean Absolute Error of less than 5 min for all cases. Finally, the predicted ramp-ups are compared with the existing fixed schedule, where significant saving windows of up to 80 min are achieved.

This study proposes a simulation-based methodology for estimating the energy saving achievable through the implementation of a just-in-time morning ramp-up procedure in a warehouse (equipped with a heat pump). In this methodology, the operation of the heating supply unit each day is initiated at a different time, aiming at achieving the desired setpoint upon (and not before) the expected arrival of the occupants. It requires the estimation of the ramp-up duration (the time it takes the heating system to bring the indoor temperature to the desired setpoint), which can be provided by machine learning-based models. To justify the corresponding required deployment investment, an accurate estimation of the resulting achievable energy saving is needed. Accordingly, physics-based energy behavior simulations are first performed. Next, various ML algorithms are employed to estimate the ramp-up duration using the simulated time-series data of indoor temperature, setpoints, and weather conditions. It is shown that the proposed pipelines can estimate the ramp-up duration with a mean absolute error of about 3 min in all indoor spaces. To assess the resulting potential energy saving, a re-simulation is conducted using ML-based ramp-up estimations for each day, resulting in an energy savings of approximately 10%.

A case study represented by an assisted living facility in Norway is modeled utilizing physics-based data-driven digital twin (DT) of the indoor thermal spaces with indoor temperature. Autoregressive Distributed Lag (ARDL), Machine Learning (ML), and Non-linear Autoregressive (NARX) models with timeseries and sliding-window cross-validation are compared. Results show that NARX models have the highest accuracy, with a MAPE score of 0.03%. In addition, the sliding-window enhanced the models’ accuracy and reduced the cyclical pattern for the autocorrelated values. The HVAC systems in this study case are representative of those found in Norwegian buildings, making the digital twin calibration applicable to other facilities.