An adaptive central difference Kalman filter approach for state of charge estimation by fractional order model of lithium-ion battery

The key issue of the model-based state of charge estimation approach is the accuracy of the battery model. In this paper, a fractional order model is built to simulate the electrochemistry dynamics of lithium-ion battery, whose model parameters are identified by adaptive genetic algorithm. Based on the computation simplification of central difference algorithm, an adaptive central difference Kalman filter by fractional order model is designed to estimate the state of charge. The designed approach is modelled by simulink and translated into C code, and then embedded in the battery management system for the validation by two dynamic cycles. Comparing experiments adopt two approaches, i.e. the central difference Kalman filter by fractional order model, the adaptive central difference Kalman filter by Thevenin model. Experimental results indicate that the designed approach has the better accuracy and robustness, and also show that fractional order model is more accurate than Thevenin model. With respect ot the ability to deal with noise, the robustness of the designed approach is verified by adding artificial noise. Experimental results show that the proposed approach has the best robustness to noise. Therefore, the proposed approach is a good candidate for the state of charge estimation in engineering practice.(c) 2021 Elsevier Ltd. All rights reserved.

Automated summary

This paper presents a model-based approach for state of charge estimation of batteries using a fractional order model and an adaptive central difference Kalman filter. The experimental results demonstrate that the proposed approach achieves higher accuracy and noise robustness compared to the Thevenin model.