sanglinwei / DiffCharge

DiffCharge: Generating EV Charging Scenarios via a Denoising Diffusion Model

Authors: Siyang Li, Hui Xiong, Yize Chen (HKUST-GZ)

Paper: https://ieeexplore.ieee.org/document/10418170, https://arxiv.org/abs/2308.09857, accepted to IEEE Transactions on Smart Grid

EV charging scenarios generated by our model:

Motivation: Recent proliferation of electric vehicle (EV) charging events has brought prominent stress over power grid operation. Due to the stochastic and volatile EV charging behaviors, the induced charging loads are extremely uncertain, posing modeling and control challenges for grid operators and charging management. Generating EV charging scenarios would aid via synthesizing a myriad of realistic charging scenarios. To this end, we propose a novel denoising Diffusion-based Charging scenario generation model DiffCharge, which is capable of generating a broad variety of realistic EV charging profiles with distinctive temporal properties.

Implementation of the main framework:

1. fetch ACN-data employed in our work: get_data.py
2. data preprocessing and structure: preprocess.py and dataset.py
3. denoising network: network.py
4. diffusion-driven generation: diffusion.py
5. training and inference: main.py
6. experimental configuration: options.py

Reproducation of the evaluation experiments:

1. baseline models: gmm.py, aae.py, timegan.py
2. metrics calculation: metrics.py, acf.py
3. tail clustering: tail.py
4. day-ahead charging energy bidding simulation: dam.py

Other useful tools: utils.py, data_client.py, figure.py

Questions? Contact Siyang at [e-mail].