The estimation of the state of the health (SoH) of the Li-ion batteries allows to improve the safety of the Li-ion battery operation and estimate the its operation time. The semi-empirical models are one of the simplest methods for the prediction of the loss of capacity and increasing of the internal resistance based on the known operation cycle of the battery system. Such models allow to take into account the effect of different stress factors such as temperature, depth of discharge, state of charge and C-rate current on the speed of the battery system degradation. Moreover, the effect of the different combination of the mentioned stress factors can be taken into account. The low computation complexity of the semi-empirical models allows to use a widely available software for the SoH estimation. The low accuracy and requirement of the big amount of the measured data for the model parameters adjustment can be attributed to the disadvantages of the semi-empirical models. However, the hight popularity of the Li-ion batteries and the significant research directed on the analysis of the Li-ion batteries degradation mechanisms leads to the improving situation with the availability of the required information for the model training in the literature from year to year. Currently, the required data for the SoH model training were presented by many researchers in literature for the most popular battery chemistries. The correct selection of these data from the literature allows to use them for the preliminary SoH estimation of the considered battery system operated at the known operation cycle. The selection of the data always should be done based on the battery chemistry. Further the cell type, dimensions and the manufactory company may be considered with purpose to improve accuracy of the semi-empirical models.
Text by Dr. Kirill Murashko