Abstract
For the purpose of daily battery maintenance, safety forecasting, and energy ladder use, it is crucial to accurately estimate the state of health (SOH) of Li-ion batteries. The external fiber Bragg grating (FBG) sensor is used in this work to collect the signal fluctuations of the wavelength shift in different areas during the Li-ion battery's charge-discharge cycle using a combination of feature analysis and data-driven method, and the characteristics of the wavelength shift and the charge- discharge process in different regions under 125 cycles are statistically analyzed. The reaction degree of Li-ion battery in different areas was compared by wavelength shift value. The wavelength shift factor and charge-discharge process factor related to battery SOH were extracted and trained in the NGO-BP regression model. The experimental results show that the combination of fiber Bragg grating wavelength shift factor and charge-discharge process factor can accurately estimate the SOH of Li-ion battery, and the prediction effect of three areas combination is the best, the goodness of fit R2 is 0.99958, the mean square error MSE is 2.71 × 10−4, the mean absolute error MAE is 1.3373 × 10−2, the root mean square error RMSE is 1.6462 × 10−2, and the mean absolute percentage error MAPE is 1.3679 × 10−2%.The combination of negative electrode and middle area uses fewer sensors to obtain higher prediction accuracy, and its cost performance is the highest. The utilization of fiber Bragg grating wavelength shift factor and charge-discharge process factor has essentially led to the accurate calculation of Li-ion battery SOH, which holds significance for battery health monitoring as well as the design and development of battery management systems.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have