Abstract:Electrochemical impedance spectroscopy (EIS) is an effective technique for Lithium-ion battery state of health diagnosis, and the impedance spectrum prediction by battery charging curve is expected to enable battery impedance testing during vehicle operation. However, the mechanistic relationship between charging curves and impedance spectrum remains unclear, which hinders the development as well as optimization of EIS-based prediction techniques. In this paper, we predicted the impedance spectrum by the battery charging voltage curve and optimized the input based on electrochemical mechanistic analysis and machine learning. The internal electrochemical relationships between the charging curve, incremental capacity curve, and the impedance spectrum are explored, which improves the physical interpretability for this prediction and helps define the proper partial voltage range for the input for machine learning models. Different machine learning algorithms have been adopted for the verification of the proposed framework based on the sequence-to-sequence predictions. In addition, the predictions with different partial voltage ranges, at different state of charge, and with different training data ratio are evaluated to prove the proposed method have high generalization and robustness. The experimental results show that the proper partial voltage range has high accuracy and converges to the findings of the electrochemical analysis. The predicted errors for impedance spectrum are less than 1.9 mΩ with the proper partial voltage range selected by the corelative analysis of the electrochemical reactions inside the batteries. Even with the voltage range reduced to 3.65–3.75 V, the predictions are still reliable with most RMSEs less than 4 mΩ.

EIS equivalent circuit model prediction using interpretable machine learning and parameter identification using global optimization algorithms

Equivalent Circuit Model Recognition of Electrochemical Impedance Spectroscopy via Machine Learning

Machine Learning Benchmarks for the Classification of Equivalent Circuit Models from Electrochemical Impedance Spectra

Multi-class AdaBoost ELM

Fault Diagnosis for Proton Exchange Membrane Fuel Cell Systems Via Improved Equivalent Circuit Model and Multiple Support Vector Machines

Robust and Fast Estimation of Equivalent Circuit Model from Noisy Electrochemical Impedance Spectra

A comparative study of global optimization methods for parameter identification of different equivalent circuit models for Li-ion batteries

Analysis of Electrochemical Impedance Data: Use of Deep Neural Networks

Battery Modeling: Fusing Equivalent Circuit Models with Data-Driven Surrogate Modeling

A Hybrid Genetic Expression Programming and Genetic Algorithm (GEP-GA) of Auto-modeling Electrical Equivalent Circuit for Particle Structure Measurement with Electrochemical Impedance Spectroscopy (EIS)

Fractional-Order Equivalent-Circuit Model Identification of Commercial Lithium-Ion Batteries

Prediction of Electrical Conductivity of Ionic Liquids: From COSMO-RS Derived QSPR Evaluation to Boosting Machine Learning

A comparative study of equivalent circuit models and enhanced equivalent circuit models of lithium-ion batteries with different model structures

Model identification of the Proton Exchange Membrane Fuel Cells by Extreme Learning Machine and a developed version of Arithmetic Optimization Algorithm

A guide to equivalent circuit fitting for impedance analysis and battery state estimation

Uneven Usage Battery State of Health Estimation via Fractional-Order Equivalent Circuit Model and AutoML Fusion

Estimation of lithium-ion battery electrochemical properties from equivalent circuit model parameters using machine learning

Development of Machine Learning Models for Ion-Selective Electrode Cation Sensor Design

Battery impedance spectrum prediction from partial charging voltage curve by machine learning

What is the appropriate data representation of electrochemical impedance spectroscopy in machine-learning analysis?

Online Estimation of Battery Equivalent Circuit Model Parameters Using Decoupled Least Squares Technique