Remaining Useful Life Prediction of Lithium Battery via Neural Network Ensemble
Authors:
- 26 November 2021
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 392-397
Abstract
Accurately predicting the Remaining Useful Life (RUL) of lithium-ion batteries is critical for accelerating the technology development. The neural network via data driven can avoid manual feature extraction and release the difficulty of model construction. However, diverse aging mechanisms, significant device variability and dynamic operating conditions have remained major challenges. The single neural network cannot tackle these challenges efficiently. In this paper, the RUL prediction of lithium battery via neural network ensemble is developed to improve the accuracy and generalization of the prediction model. The predicting results of multiple neural networks are considered as the base vectors, and the liner combination on the base vectors are conducted to obtain the ensemble RUL prediction of lithium-ion battery. The experiment is performed on the MIT-Stanford University lithium battery accelerated degradation dataset, and the experimental results show that the developed method has achieved better generalization and higher accuracy comparing with the other methods.Keywords
This publication has 16 references indexed in Scilit:
- Thermal runaway mechanism of lithium ion battery for electric vehicles: A reviewEnergy Storage Materials, 2018
- A selective ensemble method based on K-means method2015
- A review of lithium deposition in lithium-ion and lithium metal secondary batteriesJournal of Power Sources, 2014
- Comparison, Selection, and Parameterization of Electrical Battery Models for Automotive ApplicationsIEEE Transactions on Power Electronics, 2013
- Bagging, Boosting and Ensemble Methods2011
- Main aging mechanisms in Li ion batteriesJournal of Power Sources, 2005
- Neural network ensembles: combining multiple models for enhanced performance using a multistage approachExpert Systems, 2004
- Ensembling neural networks: Many could be better than allArtificial Intelligence, 2002
- Bagging predictorsMachine Learning, 1996
- Neural network ensemblesIEEE Transactions on Pattern Analysis and Machine Intelligence, 1990