Abstract: An accurate battery model is the basis of battery state estimation and energy management. Electrochemical models of lithium-ion batteries can describe their mechanism and have the advantages of clear physical interpretation and high accuracy.Currently, the commonly used electrochemical models are the pseudo-two-dimensional electrochemical model(P2D), the single-particle model(SPM), and the single-particle model with electrolyte dynamics(SPMe). Among them, the P2D model has the highest accuracy with the largest computational burden, while the two simplified models of the SPM and SPMe can run efficiently but suffer from the errors resulting from model simplification. To explore the error mechanism, the electrode current source, the solid-phase concentration, and the liquid-phase concentration of the SPM, SPMe, and P2D models are compared, and then the model errors of the simplified models are analyzed in terms of the equilibrium potential, overpotential, liquid-phase concentration polarization potential, liquid-phase ohmic potential drop, and model voltage. In addition, the model error path is given. The results show that the electrode current source is the origin of the model output voltage error, and the liquid-phase positive ohmic potential drop, the liquid-phase negative ohmic potential drop, and the negative equilibrium potential errors are the main sources to generate the model output voltage errors. The above results can provide a theoretical basis on designing error compensation for both the SPM and SPMe models. Finally, suggestions on how to compensate for the model output voltage error are also provided.

Key words: lithium-ion battery, pseudo-two-dimensional electrochemical model, single particle model, error analysis