rhced

Residential heating and cooling energy disaggregation (rhced)

This is accompanies code of paper “Ham, S.-W., Karava, P., Bilionis, I., and Braun, J. (2021). Scalable and practical heating and cooling energy disaggregation from smart thermostat and meter data for eco-feedback design. Submitted to xx.”. This is a simple demonstaration of heating and cooling (HC) energy disaggregation from the net energy consumption.

Please see our demo notebook link.

Folder structure

• data: sample data
• docs: demo notebook in .html format click.
• notebook: demo notebook based on the sample data.
• rhced: model code.
• outputs: model outputs.
• visualization: collection of R scripts to draw the figures presented in the paper.

Model structure

├──rhced
│   ├──__init__.py
│   ├──data_utils.py: Utility functions for data processing. Used for sample data demo. 
│   ├──misc.py: some utility functions for data processing. No needs for sample data demo.
│   ├──prediction.py: prediction module.
│   ├──training.py: training module.

required packages

The following Python packages are required.

• pymc3>=3.11
• pyarrow>=5.00
• pandas>=1.3.2
• numpy>=1.21.2
• psypy>=0.0.2

Visualization

It is a collection of R scripts to draw the figures presented in the paper.

You can open rhced.Rproj project, but the visualization functions are not part of the model.

The following R packages are required.

• tidyverse
• pathchwork
• lubridate

Generate .html file

• jupyter nbconvert --to html sample_bldg.ipynb

Copy the generated .html to ~/docs/xx.html to make the latest notebook is published at https://ecosang.github.io/rhced/sample_bldg.html

For exampe, in Window machine, copy sample_bldg.html ../docs/sample_bldg.html