Abstract:The rapid advancements in electric vehicle technology have elevated the lithium-ion battery to the forefront as the paramount energy storage solution. The battery's health tends to deteriorate gradually as it ages. Due to the inevitable physiochemical reactions that take place inside the battery, it undergoes degradation and at a certain point, it becomes unserviceable. The battery degradation can be estimated using state of health (SOH). This paper employs data-driven techniques to estimate the state of health (SOH) of a battery. To estimate health parameters, a vast quantity of data, such as voltage, current, and temperature, is gathered from the NASA Prognostics Center of Excellence. The data is resampled using the superior Fourier Resampling method and then fed to a machine-learning algorithm. In this study, SOH estimation is carried out using three different machine-learning techniques i.e. Long Short Term Memory (LSTM), Deep Neural Networks (DNN), and Gated Recurrent Unit (GRU). However, the performance and accuracy of SOH estimation using these algorithms are highly dependent on hyperparameter tuning. Therefore, the optimal hyperparameter tuning has been adopted in the present work to reduce the time and complexity of the estimation. Further, the performance of various proposed techniques has been compared against each other using different performance indices such as root mean square error (RMSE), mean absolute error (MAE), and R-square error. GRU technique proved to be excelling with RMSE of 0.003, MAE of 0.003, and R-square error of 0.004 while estimating the SOH of various samples of batteries. This detailed analysis will be helpful for users to evaluate the performance of a battery and plan for maintenance accurately and effectively with minimum downtime.

Battery State of Health Estimation from Discharge Voltage Segments Using an Artificial Neural Network

A Hybrid Data-Driven Approach for State of Health Estimation in Lithium-Ion Batteries

Machine Learning-Based Electrode-Level State of Health Estimation for NMC/Graphite Battery Cells

State of Health Estimations for Lithium-Ion Batteries Based on MSCNN

State of Health Estimation Based on the Long Short-Term Memory Network Using Incremental Capacity and Transfer Learning

Lithium-ion Battery Health Estimation Using DCNN Paralleled LSTM-Self Attention Networks

State of health and remaining useful life prediction for lithium-ion batteries based on differential thermal voltammetry and a long and short memory neural network

Lithium-ion battery SOH prediction based on VMD-PE and improved DBO optimized temporal convolutional network model

Optimally Tuned Gated Recurrent Unit Neural Network-Based State of Health Estimation Scheme for Lithium Ion Batteries

State of Health Estimation Method for Lithium-Ion Batteries Using Incremental Capacity and Long Short-Term Memory Network

A hybrid neural network based on variational mode decomposition denoising for predicting state-of-health of lithium-ion batteries

Online State of Health Estimation with Deep Learning Frameworks Based on Short and Random Battery Charging Data Segments

Industrial Battery State-of-Health Estimation with Incomplete Limited Data Towards Second-Life Applications

A New Hybrid Neural Network Method for State-of-Health Estimation of Lithium-Ion Battery

Estimation of the state of health (SOH) of batteries using discrete curvature feature extraction

Battery State-of-Health Estimation: A Step towards Battery Digital Twins

A Comparative Study of Battery State-of-health Estimation Based on Empirical Mode Decomposition and Neural Network

Prediction of Lithium Battery Health State Based on Temperature Rate of Change and Incremental Capacity Change

Robust estimation of lithium-ion battery state of health based on electro-thermal features and machine learning

State-of-Health Prediction of Lithium-Ion Batteries Based on CNN-BiLSTM-AM

Data-based health indicator extraction for battery SOH estimation via deep learning