Commit a9007010 authored by Avisek Naug's avatar Avisek Naug
Browse files

update readme

parent 4f0a464c
......@@ -47,19 +47,25 @@ The repository contains the following files:
* `agent.py`: Defines the `agent` and functions to train and test the agent.
* `HVAC_environment.py`: Defines the environment representing the HVAC system.
* `ah_api.py`: Pulls the weekly building + solar data for relearning.
* `moderetrain.py`: used to retrain the LSTM model on the new week data.
* `helperfunctions.py`: set of helper methods for reading and processing raw historical and relearning data
* `Building_Environment_RL_Demo.ipynb`: A jupyter notebook demoing the training and testing process+new test processes added
* `controller.py`: The main script that will run the control functionality.
* `agent_weights_actor.h5f`, `agent_weights_critic.h5f`: The initial control policies to use.
* `solar_irradiance.py`: Pulls solar irradiation data using Solcast API.
* `*.slurm`: Files used to train for the best weights
* `LSTMstateful_batchsize1.h5`: LSTM model for energy prediction. Scaled data in -> scaled data out
* `GradBoostReg_scaled.joblib`:GBR model for energy prediction. Scaled data in -> scaled data out
* `GradBoostReg_nonscaled.joblib`: GBR model for energy prediction. non-Scaled data in -> non-scaled data out
* `Summer_2019_5min.pkl`:Summer 2019 data base. Format of the data shown below
* `*.slurm`: Files used for offline weekly training
* `weights.best.hdf5`: LSTM model for energy prediction. Scaled data in -> scaled data out
* `RL_relearn_data.pkl`:One year plus worth of historical data going back from 22nd October 2019. Format of the data shown below
| Dates | AirTemp (F) | AirHum (%)| Ghi (W/m2)| DischTemp (F)| TotalE (kBtus)|
| `Dates` | `OAT` (F) | `OAH` (%) | `Ghi` (W/m2)| `SAT` (F) | `TotalE` (kJ) |
|---------------------|-------------|-----------|-----------|---------------|---------------|
| 2019-05-01 00:10:00 | 73.28 | 71.87 | 53.753710 | 28.0 | 35.551946 |
* `OAT`: Outside air temperature
* `OAH`: Outside air humidity
* `Ghi`: Global solar irradiation at Nashville
* `SAT`: Temperature of air coming out of the Air Handling Unit
* `TotalE`: Total Chilled Water Energy and Hot Water energy circulating through the entire building
To periodically train on data from the Alumni Hall under the server implementation, replace the datapath string appropriately.
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment