Transformer Network for Remaining Useful Life Prediction of Lithium-Ion Batteries

同类别比较
名称	排名	指标
circuitnet/CircuitNet	17,967	语言: Python 星标数: 515 分叉数: 86 简介: CircuitNet: An Open-Source Dataset for Machine Learning Applications in Electronic Design Automation (EDA)
clvrai/SSGAN-Tensorflow	17,968	语言: Python 星标数: 515 分叉数: 164 简介: A Tensorflow implementation of Semi-supervised Learning Generative Adversarial Networks (NIPS 2016: Improved Techniques for Training GANs).
Lizhen0628/text_classification	17,985	语言: Python 星标数: 514 分叉数: 73 简介: 使用rnn,lstm,gru,fasttext,textcnn,dpcnn,rnn-att,lstm-att,兼容huggleface/transformers，以及以transforemrs作为词嵌入模型，后面接入cnn、rnn、attention等等做文本分类。以及各个模型的对比
VCIP-RGBD/DFormer	17,988	语言: Python 星标数: 514 分叉数: 63 简介: DFormers for Multimodal Semantic Segmentation
alteryx/compose	17,992	语言: Python 星标数: 514 分叉数: 51 简介: A machine learning tool for automated prediction engineering. It allows you to easily structure prediction problems and generate labels for supervised learning.
chaiyujin/glow-pytorch	17,995	语言: Python 星标数: 514 分叉数: 81 简介: pytorch implementation of openai paper "Glow: Generative Flow with Invertible 1×1 Convolutions"
google-deepmind/dqn_zoo	18,000	语言: Python 星标数: 514 分叉数: 86 简介: DQN Zoo is a collection of reference implementations of reinforcement learning agents developed at DeepMind based on the Deep Q-Network (DQN) agent.
koaning/doubtlab	18,002	语言: Python 星标数: 514 分叉数: 17 简介: Doubt your data, find bad labels.

名称

排名

指标

circuitnet/CircuitNet

17,967

语言:: Python
星标数:: 515
分叉数:: 86
简介:: CircuitNet: An Open-Source Dataset for Machine Learning Applications in Electronic Design Automation (EDA)

clvrai/SSGAN-Tensorflow

17,968

语言:: Python
星标数:: 515
分叉数:: 164
简介:: A Tensorflow implementation of Semi-supervised Learning Generative Adversarial Networks (NIPS 2016: Improved Techniques for Training GANs).

Lizhen0628/text_classification

17,985

语言:: Python
星标数:: 514
分叉数:: 73
简介:: 使用rnn,lstm,gru,fasttext,textcnn,dpcnn,rnn-att,lstm-att,兼容huggleface/transformers，以及以transforemrs作为词嵌入模型，后面接入cnn、rnn、attention等等做文本分类。以及各个模型的对比

VCIP-RGBD/DFormer

17,988

语言:: Python
星标数:: 514
分叉数:: 63
简介:: DFormers for Multimodal Semantic Segmentation

alteryx/compose

17,992

语言:: Python
星标数:: 514
分叉数:: 51
简介:: A machine learning tool for automated prediction engineering. It allows you to easily structure prediction problems and generate labels for supervised learning.

chaiyujin/glow-pytorch

17,995

语言:: Python
星标数:: 514
分叉数:: 81
简介:: pytorch implementation of openai paper "Glow: Generative Flow with Invertible 1×1 Convolutions"

google-deepmind/dqn_zoo

18,000

语言:: Python
星标数:: 514
分叉数:: 86
简介:: DQN Zoo is a collection of reference implementations of reinforcement learning agents developed at DeepMind based on the Deep Q-Network (DQN) agent.

koaning/doubtlab

18,002

语言:: Python
星标数:: 514
分叉数:: 17
简介:: Doubt your data, find bad labels.

关键参数与规格
参数	值
语言	Jupyter Notebook
类别	Model Development
子类别	Domain-specific training
关键词	Timeseries
星标数	514
分叉数	83
贡献者	1
创建日期	2021-06-21
更新日期	2026-10-14
1 天星标数	0
7 天星标数	1
排名	17,990
上次排名	17,762
排名变化	-228
7 天星标排名	11,557
上次 7 天星标排名	11,604
7 天星标排名变化	47
总星标排名	18,957
上次总星标排名	18,928
总星标排名变化	-29
国家	China

参数

值

语言

Jupyter Notebook

类别

Model Development

子类别

Domain-specific training

关键词

Timeseries

星标数

514

分叉数

贡献者

创建日期

2021-06-21

更新日期

2026-10-14

1 天星标数

7 天星标数

排名

17,990

上次排名

17,762

排名变化

-228

7 天星标排名

11,557

上次 7 天星标排名

11,604

7 天星标排名变化

总星标排名

18,957

上次总星标排名

18,928

总星标排名变化

-29

国家

China

Figure

Results of Transformer

Figures of NASA

Figures of CALCE

Supplement

Due to the length of the paper, the two parameters of dropout and noise_level are not discussed. By setting these two parameters, better results can be obtained than in the paper.

noise level = 0.01: Setting the value of 1% disturbance is best: too large will degrade performance, too small will have little effect.
dropout = 1e-4~1e-3: Set a small value for the network dropout to ensure the robustness of the model.

Packages

pytorch 1.8.0
pandas 0.24.2
mixture_of_experts 0.2.1 (for AttMoE, github: https://github.com/lucidrains/mixture-of-experts)

Update

6/5/2024, add figures of model and prediction
1/3/2024, upload the open sorce of AttMoE
24/2/2022，Change some variable names

Dataset CALCE processing reference

https://github.com/konkon3249/BatteryLifePrediction

E-mail

Please feel free to contact me: zhouxiuze@foxmail.com

More （更多内容）

马里兰大学锂电池数据集 CALCE，基于 Python 的锂电池寿命预测: https://snailwish.com/437/
NASA 锂电池数据集，基于 Python 的锂电池寿命预测: https://snailwish.com/395/
NASA 锂电池数据集，基于 python 的 MLP 锂电池寿命预测: https://snailwish.com/427/
NASA 和 CALCE 锂电池数据集，基于 Pytorch 的 RNN、LSTM、GRU 寿命预测: https://snailwish.com/497/
基于 Pytorch 的 Transformer 锂电池寿命预测: https://snailwish.com/555/
锂电池研究之七——基于 Pytorch 的高斯函数拟合时间序列数据: https://snailwish.com/576/

Citation

@article{chen2022transformer,
  title={Transformer network for remaining useful life prediction of lithium-ion batteries},
  author={Chen, Daoquan and Hong, Weicong and Zhou, Xiuze},
  journal={Ieee Access},
  volume={10},
  pages={19621--19628},
  year={2022},
  publisher={IEEE}
}

@article{chen2024attmoe,
  title={AttMoE: Attention with Mixture of Experts for remaining useful life prediction of lithium-ion batteries},
  author={Chen, Daoquan and Zhou, Xiuze},
  journal={Journal of Energy Storage},
  volume={84},
  pages={110780},
  year={2024},
  publisher={Elsevier}
}

RUL

记录了多少星标和分叉？

仓库排名是多少？

记录的语言和类别是什么？

哪个仓库排名最接近？