Electrochemical impedance characteristics at various conditions for commercial solid-liquid electrolyte lithium-ion batteries: Part. 2. Modeling and prediction

Solid-liquid electrolyte lithium-ion batteries (SLELBs) have good commercial viability in electric vehicle applications because they combine the safety of solid electrolyte lithium-ion batteries with the high ionic conductivity of liquid electrolyte lithium-ion batteries (LELBs). The safe and efficient operation of electric vehicles is inseparable from the key battery management algorithms such as battery state of charge (SOC), state of health and state of power estimation. In the process of designing battery management algorithms for SLELBs, it is essential to have an accurate understanding of battery behavior under different influencing factors and to build a high-fidelity battery simulation model. Electrochemical impedance spectroscopy (EIS) can be used to study the electrode process dynamics and ion transport mechanism in lithium-ion batteries. It is an urgent challenge to use EIS to experiment and analyze the characteristic impedances of SLELBs under the full-scale factors and to construct the battery model and simulate the battery impedance under the premise of a reasonable number of tests.In this series of two papers, we first propose a regression model of temperature and SOC One-Way factor SLELB characteristic impedances through a full-scale experimental design under the theoretical framework of experimental optimization design and statistical analysis. Then, in order to construct the model under the premise of a reasonable number of tests, an orthogonal experiment is designed and an orthogonal piecewise polynomial Arrhenius-simplified equivalent circuit model (SECM) for SLELB impedance prediction of the Two-Way factor of temperature and SOC is established. These models can accurately describe SLELB behavior under different factors, improve the understanding of SLELB electrode processes, and predict the impedance spectrum of SLELB, all of which support the key battery management algorithms. In the first paper, the characteristic impedances in the EIS were identified to facilitate a systematic analysis at different temperatures and SOCs in a full-scale experimental design. The Arrhenius and polynomial regression models of the One-Way factor SLELB characteristic impedances of temperature and SOC were established.In this second paper, an equivalent circuit model (ECM) based on the first principles of the electrode process of SLELB is established and simplified according to the EIS impedance frequency range to obtain the SECM. In order to construct the battery model under the premise of a reasonable number of experiments, we design an orthogonal experiments scheme to reduce the number of experiments to half that of the full-scale experiments. An orthogonal piecewise polynomial Arrhenius (OPPA) temperature and SOC Two-Way factor parameter model is established using ANOVA considering the interaction of the ECM model parameters. The experimental results show that OPPA has an accurate prediction performance for ECM parameters with a reasonable number of segments and orders. The predictive performance of OPPA-SECM is similar to that of full-scale piecewise polynomial Arrhenius (FPPA)-SECM, and the number of OPPA-SECM experiments is reduced by half that of FPPA-SECM experiments.(c) 2021 Elsevier Ltd. All rights reserved.

Automated summary

Solid-liquid electrolyte lithium-ion batteries (SLELBs) have commercial viability in electric vehicle applications. It is important to accurately understand battery behavior under different factors and build simulation models. Electrochemical impedance spectroscopy (EIS) can be used to study electrode processes and ion transport mechanisms in lithium-ion batteries. In this series of papers, regression models and an orthogonal experiment design are proposed to predict SLELB impedance under temperature and SOC factors. An equivalent circuit model based on the electrode process is established and simplified to obtain the simplified equivalent circuit model (SECM). The experimental results show that the SECM has accurate prediction performance and reduces the number of required experiments.