AI Baseline Energy Model for Building Savings Accuracy

The patent segments the energy modeling process into two distinct approaches: detailed facility models for high-precision applications and statistical correlation models for rapid assessments. This segmentation allows users to select the appropriate level of detail based on specific needs, reducing unnecessary complexity while maintaining accuracy where required.

The patent introduces statistical correlation models as an intermediary solution between simple degree-day models and complex detailed facility models. These intermediate models use readily available data (monthly degree-days and energy bills) combined with statistical regression to provide reasonably accurate energy savings estimates without requiring extensive on-site evaluation, thus mediating the trade-off between precision and complexity.

The patent applies preliminary action by preparing statistical correlation models in advance using historical data and statistical regression analysis. These pre-developed models can be quickly deployed to estimate energy savings without requiring time-consuming on-site evaluations at the time of implementation, thus reducing loss of time while maintaining reasonable accuracy.

The patent uses copying by creating simplified statistical representations (correlation models) that replicate the essential behavior of complex detailed facility models. These copied models use monthly degree-days and energy bill data to approximate the results of full detailed models, providing accurate enough estimates for portfolio-level assessments without requiring the original complex models to be executed for each building.