*4.4. SOC Estimators—Real Time Implementation*

As was mentioned in the previous section, due to their predictor–corrector structure, each SOC estimator becomes a recursive algorithm, more straightforward to implement in real-time and very efficient in terms of computation. Both Li-ion battery models are also simple, easy to design and quickly deploy, especially the Simscape battery model based directly on the manufacturer's battery specifications. Besides, MATLAB-Simulink software platform provides a valuable and practical Simscape/SimPower Systems library, helpful for use in design and implementation of different HEV and EV powertrain configurations.

### *4.5. SOC Estimator Robustness Performance Analysis—Statistical Criteria*

The values of statistical criteria from Table 1 provide the SOC accuracy of both battery models concerning ADVISOR estimate, beneficial for Li-ion battery model validation performed for an FTP-75 driving cycle profile test. The statistical criteria values from Tables A1–A4 are valuable for analysing the SOC robustness performance of all three SOC estimators. Based on the information extracted from Tables A1–A4 for each SOC estimator, it seems that AEKF SOC is more robust compared to the other two SOC estimators, as is quite evident for the Simulink model. Unfortunately, it is diffficult to make a complete performance analysis by comparison of the results obtained by similar SOC estimators reported in the literature. This happens since many researchers use different input current profiles and various statistical criteria that do not match with those used in our research. However, for the cases that match with our driving cycle profile test, the information collected in Tables 1 and A1, Tables A2–A4 can be useful for analysing all similar situations. Thus, the present research work can be a valuable source of inspiration for readers and researchers.
