Non-destructive state-of-health diagnosis algorithm for blended electrode lithium-ion battery

Optimisation methods based on half-cell measurements provide efficient non-destructive aging diagnosis for lithium-ion batteries. However, a blend electrode using this approach could bias the observations and lead to false aging scenario determination. The present study shows a non-intrusive method to quantify both the state of health of a cell and the partial aging of a blend active material LMFP:NCA. From the classical optimisation of the half-cell positions on a cell pseudo-open-circuit voltage, a blend submodel is added to integrate the underlying changes into the blend mass fraction. After the optimisation was performed on the battery check-up measurements, the aging phenomena were gathered into degradation modes that were quantified throughout the cell lifetime, and the changes in the electrode positions were converted into losses of lithium inventory, losses of positive and negative active materials, and an increase in ohmic resistance. The partial aging of the blend components was calculated using the mass fraction evolution of the corresponding loss of the electrode. Investigations were conducted on a 30-Ah high-power LMFP:NCA/graphite lithium-ion prototype battery. The basic root mean square error optimisation criterion was associated with differential methods (incremental capacity and differential voltage) to validate the numerical results and enhance the robustness of the optimisation.