Attention-based-Synthetic-Battery-Data-Augmentation-Technique

Get started

This project has been executed via google colab.

Steps To Follow:

1. Download the code as a zip file and extract the contents.

2. Create a new folder in google drive and upload the extracted contents into it.

3. The code to be executed in google colab:

4. Make the necessary changes mentioned in the code to get the result.

5. The datsasets used can be found in the informer_model/Dataset-main/datasets folder. Any new dataset to be used should be uploaded in the mentioned folder.

6. All changes pertaining to prediction length, dataset etc can be done in main_informer.py script.

7. The dataset can be changed in main_informer.py in data parser under Battery (The name of the csv file used should be mentioned under the data parameter. The target parameter can be varied under the parameter T).

Usage

To train and evaluate the Informer model on a dataset the following command is used in the google colab notebook :

!python -u main_informer.py --model informer --data Battery --train_epochs 4 --attn prob --freq s --features S 

The detailed descriptions about the arguments are as following:

Parameter name	Description of parameter
model	The model used in the experiment. This can be set to informer, informerstack
data	The dataset name
train_epochs	The number of epochs the code has to run for (defaults to 6)
attn	Attention used in encoder (defaults to prob). This can be set to prob (informer), full (transformer)
freq	Freq for time features encoding (defaults to h). This can be set to s,t,h,d,b,w,m (s:secondly, t:minutely, h:hourly, d:daily, b:business days, w:weekly, m:monthly
features	The forecasting task (defaults to M). This can be set to M,S,MS (M : multivariate predict multivariate, S : univariate predict univariate, MS : multivariate predict univariate)

On running the main code a results folder will be created where the prediction output is stored.

Acknowledgement

We appreciate the following github repos a lot for their valuable code base or datasets:

https://github.com/zhouhaoyi/Informer2020

Diao, W., Saxena, S., Pecht, M. Accelerated Cycle Life Testing and Capacity Degradation Modeling of LiCoO2 -graphite Cells. J. Power Sources 2019, 435, 226830.

https://web.calce.umd.edu/batteries/data.htm