Leveraging Massive Datasets to Close the Loop on Building Performance Goals
The project aims to optimize energy consumption, maintain occupant comfort, and monitor energy systems in UCI buildings by developing toolsets for performance review and malfunction detection.
Team Members:
Hsiang Wei Wu (Team Lead), Yifei Ren, and Rong Hu
Project Deliverables:
- Building Energy Performance Review Toolset
- Assessed two-factor energy consumption prediction model and LBNL energy consumption prediction model
- Investigated the offline and online change point detection to detect changes in operation and suggest a training period
- Correlation-Based Malfunction Detection Toolset for Building Retro-Commissioning
- Utilized correlation analysis to identify abnormal units
- Applied a machine learning (ML) model to discover potential factors and create visualizations
