Generation and representation of synthetic smart meter data

Abstract

Advanced energy algorithms running at big-data scale will be necessary to identify, realize, and verify energy savings to meet government and utility goals of building energy efficiency. Any algorithm must be well characterized and validated before it is trusted to run at these scales. Smart meter data from real buildings will ultimately be required for the development, testing, and validation of these energy algorithms and processes. However, for initial development and testing, smart meter data are difficult to work with due to privacy restrictions, noise from unknown sources, data accessibility, and other concerns which can complicate algorithm development and validation. This study describes a new methodology to generate synthetic smart meter data of electricity use in buildings using detailed building energy modeling, which aims to capture the variability and stochastics of real energy use in buildings. The methodology can create datasets tailored to represent specific scenarios with known truth and controllable amounts of synthetic noise. Knowledge of ground truth also allows the development and validation of enhanced processes which leverage building metadata, such as building type or size (floor area), in addition to smart meter data. The methodology described in this paper includes the key influencing factors of real-world building energy use including weather data, occupant-driven loads, building operation and maintenance practices, and special events. Data formats to support workflows leveraging both synthetic meter data and associated metadata are proposed and discussed. Finally, example use cases of the synthetic meter data are described to illustrate potential applications.

A synthetic building operation dataset

“…Then, we modified the seed models to represent three energy efficiency levels by changing the building envelope properties, lighting, MELs, and HVAC system efficiencies. We then modified the schedules for zone-level occupancy, lighting, MELs, and thermostat setpoint, to reflect more realistic building operations 17 . Next, we ran simulations with the updated models, which utilized thirty years’ historical weather data plus a Typical Meteorological Year (TMY3) 22 weather data.…”

“…For grey-box and data-driven approaches, such a comprehensive dataset is critical for training reliable models. As of now, there are numerous efforts in either collecting data from measurements 13 – 16 or synthesizing data with simulations 17 , 18 . However, each of the dataset has its strengths and limitations.…”

A Survey on the Use of Synthetic Data for Enhancing Key Aspects of Trustworthy AI in the Energy Domain: Challenges and Opportunities

“…For example, ML models can be trained on pure real-world temperature or electricity time series data, which in turn, generate synthetic data [97][98][99][100].…”

Occupant-Centric key performance indicators to inform building design and operations

“…In actual buildings, occupant movements and interactions with building systems are highly dynamic and stochastic. Hong et al (Hong, Macumber, et al 2020) proposed a methodology to simulate those dynamics.…”