Abstract:Predicting the evolution of multiphysics battery systems face severe challenges, including various aging mechanisms, cell-to-cell variation and dynamic operating conditions. Despite significant progress, solving real-life battery problems with noisy and missing data and high-dimensional parameter space are either difficult or impossible. In this paper, we design and evaluate feature-based machine learning techniques for estimating the capacity of large format LiFePO4 batteries in EV applications and hence predicting the trajectory of capacity fade based on the estimations. To probe the feature space, we generate a comprehensive dataset consisting of 420 cells and 9 battery packs (178 cells in-series for each one) with more than 10,000 validation data derived from the cloud platform. A two-step noise reduction method is applied to de-noise the scattered field data (30s sampling interval). Totally, 39 domain features are engineered using the reconstructed segments of battery charging data based on the differential methods (increment capacity and differential voltage), which steer the learning process towards accurate and physically consistent predictions by leveraging the stacking ensemble learning. The stacking, comprised of four base learning models referred to as level-1 predictors and a meta-learner referred to as level-2 predictor applied to combine predictions of base learners with probability distributions using an extended set of meta-level features offers exciting opportunities for better accuracy and improved generalization. Our best models achieve 0.28% mean absolute percentage error (MAPE) and 0.55% root mean squared percent error (RMSPE) for battery capacity estimation. Further, 1.22% average percentage error is achieved in the prediction of remaining useful lifetime (RUL) under different conditions of driving distance (km) and service time (day) by building capacity fade trajectory based on a Bayesian regression. This work highlights the promise of machine learning modelling using domain-specific features for accurate estimation and prediction of real-life battery systems based on field data collection.

Ultra-early prediction of lithium-ion battery performance using mechanism and data-driven fusion model

A Physics-Informed Hybrid Multitask Learning for Lithium-Ion Battery Full-Life Aging Estimation at Early Lifetime

A data and physical model joint driven method for lithium-ion battery remaining useful life prediction under complex dynamic conditions

An Empirical-informed Model for the Early Degradation Trajectory Prediction of Lithium-ion Battery

A generalizable, data-driven online approach to forecast capacity degradation trajectory of lithium batteries

Predicting Capacity Fading Behaviors of Lithium Ion Batteries: an Electrochemical Protocol-Integrated Digital-Twin Solution

Deep Learning to Predict Battery Voltage Behavior after Uncertain Cycling-Induced Degradation

Early prediction of remaining useful life for Lithium-ion batteries based on a hybrid machine learning method

Early prediction of lithium-ion battery lifetime via a hybrid deep learning model

A novel hybrid model for lithium-ion batteries lifespan prediction with high accuracy and interpretability

Fast Capacity Prediction of Lithium-Ion Batteries Using Aging Mechanism-Informed Bidirectional Long Short-Term Memory Network.

Hybrid Fusion for Battery Degradation Diagnostics Using Minimal Real-World Data: Bridging Laboratory and Practical Applications

Predict the lifetime of lithium-ion batteries using early cycles: A review

An End-to-End Learning Framework for Early Prediction of Battery Capacity Trajectory

Machine learning for predicting battery capacity for electric vehicles

Data-Driven Prediction of Li-Ion Battery Degradation Using Predicted Features

Data-driven prediction of battery failure for electric vehicles

A Data-Driven Method With Mode Decomposition Mechanism for Remaining Useful Life Prediction of Lithium-Ion Batteries

An Integrated Method of the Future Capacity and RUL Prediction for Lithium-Ion Battery Pack

Deep Learning Powered Lifetime Prediction for Lithium-Ion Batteries Based on Small Amounts of Charging Cycles