skip to main content
US FlagAn official website of the United States government
dot gov icon
Official websites use .gov
A .gov website belongs to an official government organization in the United States.
https lock icon
Secure .gov websites use HTTPS
A lock ( lock ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.

Explore Research Products in the PAR
It may take a few hours for recently added research products to appear in PAR search results.


  • NSF Public Access
  • Search Results
  • Uncertainty Matters: Bayesian Probabilistic Forecasting for Residential Smart Meter Prediction, Segmentation, and Behavioral Measurement and Verification
Title: Uncertainty Matters: Bayesian Probabilistic Forecasting for Residential Smart Meter Prediction, Segmentation, and Behavioral Measurement and Verification
As new grid edge technologies emerge—such as rooftop solar panels, battery storage, and controllable water heaters—quantifying the uncertainties of building load forecasts is becoming more critical. The recent adoption of smart meter infrastructures provided new granular data streams, largely unavailable just ten years ago, that can be utilized to better forecast building-level demand. This paper uses Bayesian Structural Time Series for probabilistic load forecasting at the residential building level to capture uncertainties in forecasting. We use sub-hourly electrical submeter data from 120 residential apartments in Singapore that were part of a behavioral intervention study. The proposed model addresses several fundamental limitations through its flexibility to handle univariate and multivariate scenarios, perform feature selection, and include either static or dynamic effects, as well as its inherent applicability for measurement and verification. We highlight the benefits of this process in three main application areas: (1) Probabilistic Load Forecasting for Apartment-Level Hourly Loads; (2) Submeter Load Forecasting and Segmentation; (3) Measurement and Verification for Behavioral Demand Response. Results show the model achieves a similar performance to ARIMA, another popular time series model, when predicting individual apartment loads, and superior performance when predicting aggregate loads. Furthermore, we show that the model robustly captures uncertainties in the forecasts while providing interpretable results, indicating the importance of, for example, temperature data in its predictions. Finally, our estimates for a behavioral demand response program indicate that it achieved energy savings; however, the confidence interval provided by the probabilistic model is wide. Overall, this probabilistic forecasting model accurately measures uncertainties in forecasts and provides interpretable results that can support building managers and policymakers with the goal of reducing energy use.  more » « less
Award ID(s):
1941695
PAR ID:
10297423
Author(s) / Creator(s):
; ; ;
Date Published:
Journal Name:
Energies
Volume:
14
Issue:
5
ISSN:
1996-1073
Page Range / eLocation ID:
1481
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. ; ; (, ICASP14 - 14th International Conference on Applications of Statistics and Probability in Civil Engineering)
    Hot temperatures drive excessive energy use for space-cooling in built environments. In a building, a system operator could save costs by making better decisions under the uncertainties associated with urban temperature and future energy demands. In this paper, we assess the impact of urban weather modeling on energy cost, using a value of information (VoI) analysis, in a day-ahead (DA) electricity market. To do that, we combine two probabilistic models: (a) a model for forecasting urban temperature and (b) a model for forecasting hourly net electric load of a building given ambient urban temperature. We then quantify the impact of better urban weather modeling by propagating the uncertainty from the temperature model to the load forecasting model. We perform a numerical case study on residential building prototypes located in the city of Pittsburgh. The result indicates that using a better weather model could save 4.34-8.22% of the electricity costs for space-cooling. 
    more » « less
  2. ; (, Energies)
    Accurate long-term electricity load forecasting is critical for energy planning, infrastructure development, and risk management, especially under increasing uncertainty from climate and economic shifts. This study proposes a multi-resolution probabilistic load forecasting framework that leverages temporal hierarchies to generate coherent forecasts at hourly, daily, monthly, and yearly levels. The model integrates climate and economic indicators and employs tailored forecasting techniques at each resolution, including XGBoost and ARIMAX. Initially incoherent forecasts across time scales are reconciled using advanced methods such as Ordinary Least Squares (OLS), Weighted Least Squares with Series Variance Scaling (WLS_V), and Structural Scaling (WLS_S) to ensure consistency. Using historical data from Alberta, Canada, the proposed approach improves the accuracy of deterministic forecasts and enhances the reliability of probabilistic forecasts, particularly when using the OLS reconciliation method. These results highlight the value of temporal hierarchy structures in producing high-resolution long-horizon load forecasts, providing actionable insights for utilities and policymakers involved in long-term energy planning and system optimization. 
    more » « less
  3. ; ; ; ; ; (, Zenodo)
    Summary:   This repository contains the datasets used in the 2025 PG&E Energy Analytics Challenge on Electricity Load Forecasting, co-organized by the IISE Energy Systems Division and the INFORMS Quality, Statistics, and Reliability Section, and graciously sponsored by the Pacific Gas & Electric Company (PG&E).  The repository contains two files: Train.xlsx: Contains three years of hourly electric loads from San Diego, California (Years 2020-2022), as well as exogenous weather information at five neighboring sites within the San Diego area.    Test.xlsx: Contains one year of hourly electric loads from San Diego, California (Year 2023), as as well as exogenous weather information at five neighboring sites within the San Diego area.    Background:  This competition aimed to predict electricity loads for a specific location within the CAISO system. Accurate load forecasting is critical for managing electricity distribution within California’s diverse and dynamic energy market. Load patterns can vary significantly due to factors such as weather conditions, local supply and demand, and the mix of nearby energy generation sources.   During the competition, the specific location and the actual years were not disclosed to the participants. Participants were then asked to generate a year’s worth of load forecasts using historical load values and exogeneous weather information. Predictor data were not allowed to be taken from future time periods: When predicting the load values at a particular day, the model was allowed to only use predictor variable information from that time period or before. For example, if a team is predicting the load for Day 5 Hour 3 in Year 3, the model can only take in predictor values from Day 5 Hour 3 in Year 3, or before. Furthermore, during the competition, the load information for the test set (highlighted in yellow in the test data file) was reserved from all participants, who were asked to submit their predictions for the full year.    This is the second offering of the PG&E Energy Analytics challenge, complementing the first offering in 2024 which focused on electricity price forecasting [Results of the 2024 challenge: Aziz Ezzat, A., Mansouri, M., Yildirim, M., & Fang, X. (2025). IISE PG&E Energy Analytics Challenge 2024: Forecasting day-ahead electricity prices. IISE Transactions, 1–13. https://doi.org/10.1080/24725854.2024.2447049].  
    more » « less
  4. ; ; (, 2021 IEEE Power & Energy Society General Meeting (PESGM))
    null (Ed.)
    A framework for the generation of synthetic time-series transmission-level load data is presented. Conditional generative adversarial networks are used to learn the patterns of a real dataset of hourly-sampled week-long load profiles and generate unique synthetic profiles on demand, based on the season and type of load required. Extensive testing of the generative model is performed to verify that the synthetic data fully captures the characteristics of real loads and that it can be used for downstream power system and/or machine learning applications. 
    more » « less
  5. ; ; ; (, ASME Journal of Engineering for Sustainable Buildings and Cities)
    Abstract Accurately forecasting electric vehicle (EV) charging demand is critical for managing peak loads and ensuring grid stability in regions with increasing EV adoption. Residential household peak energy usage and EV charging patterns vary significantly across areas, influenced by geographic accessibility, sociodemographic factors, charging preferences, and EV attributes. Averaging data across regions can overlook these differences, leading to an underestimation of charging demand disparities and risking grid overload during peak periods. This study introduces a spatiotemporal trip chain-based EV charging schedule simulation method to address these challenges. The methodology integrates sociodemographic and geographic data with the large language model to generate trip chains, which serve as the basis for simulating EV charging schedules and aggregating regional energy loads to forecast peak demand. A case study of Pescadero, CA employs synthetic profiles, derived from Census statistics, to model local households as EV owners and validate the practical applicability of this approach. The results emphasize the representativeness of the trip chain generation model and the effectiveness of the EV charging schedule simulation model in accurately forecasting energy consumption patterns and assessing peak load impacts. By combining sociodemographic and geographic insights, this study provides a robust tool for evaluating the peak load impacts of EV charging. 
    more » « less
  • Citation Formats
  • MLA
  • APA
  • Chicago
  • BibTeX
  • Save / Share this Record
  • Send to Email