State‐of‐health estimation and remaining useful life for lithium‐ion battery based on deep learning with Bayesian hyperparameter optimization

2021 ◽  
Author(s):  
Depeng Kong ◽  
Shuhui Wang ◽  
Ping Ping
Keyword(s):  
Deep Learning ◽  
Lithium Ion Battery ◽  
Lithium Ion ◽  
Remaining Useful Life ◽  
State Of Health ◽  
Hyperparameter Optimization ◽  
Useful Life

Electrochemical Model-Based Aging Characterization of Lithium-Ion Battery Cell in Electrified Vehicles

2018 ◽  
Author(s):  
Meng Huang ◽  
Mrinal Kumar
Keyword(s):  
Lithium Ion Battery ◽  
Lithium Ion ◽  
Nearest Neighbor Search ◽  
Remaining Useful Life ◽  
Physical Parameters ◽  
Capacity Fade ◽  
State Of Health ◽  
Useful Life ◽  
Aging Parameter

Lithium-ion battery cycle- and calendar-life remain to be one of the greatest uncertainties for the advanced energy storage systems. Accurate characterization of battery aging has been crucial for battery state-of-health (SOH) estimation and the remaining useful life (RUL) prediction. The formation-and-growth of the solid-electrolyte interphase (SEI) has been widely recognized as one of the most prominent battery degradation mechanisms. It consumes the cyclable lithium within the cell and ultimately leads to the capacity fade which cannot be measured directly onboard. This study evaluates the multi-scale multi-physics battery models and their respective aging mechanisms as well as the corresponding characterization metrics. Then the reduced order single particle model (SPM) is selected in this study, given its parametric dependence on both electrochemical and physical parameters as well as its compatibility to the available measurements in vehicle for aging characterization. nLi, the total moles of cyclable lithium within the cell, is identified as a valid aging parameter that can effectively characterize the capacity fade through the interpretation of experimental aging data. This study also investigates into the potentially optimal testing profile and the sufficient amount of data required for the accurate aging characterization. Then the method of brute force nearest neighbor search (NNS) is applied to derive the long-term evolution trend of the aging parameter nLi, which can serve as a key benchmark for validating the in-vehicle implementable algorithms for battery state-of-health (SOH) estimation and as an important foundation for predicting the remaining useful life (RUL) of battery.

scholarly journals Remaining Useful Life Prediction for Lithium-Ion Battery: A Deep Learning Approach

IEEE Access ◽  
2018 ◽  
Vol 6 ◽  
pp. 50587-50598 ◽  
Author(s):  
Lei Ren ◽  
Li Zhao ◽  
Sheng Hong ◽  
Shiqiang Zhao ◽  
Hao Wang ◽  
...  
Keyword(s):  
Deep Learning ◽  
Lithium Ion Battery ◽  
Life Prediction ◽  
Lithium Ion ◽  
Remaining Useful Life ◽  
Learning Approach ◽  
Useful Life
Sign in / Sign up

Export Citation Format

Share Document