This paper considers the estimation of the state-of-charge of rechargeable batteries based on a classifier trained using two methods. One method uses the values of the parameters in an equivalent circuit model, identified using a frequency-domain approach. The other method is based on a mathematical approximation of the battery voltage time-response to a given 3 s current signal. Classification resorts to a linear discriminant analysis classifier trained both by experimental data and by data obtained through augmentation methods. It is shown that the time-domain based classifier may achieve better performance in terms of probability of correct state-of-charge classification, using experiments of significant less duration than those associated with the usage of the frequency-domain experiments.
Time-domain Battery State-of-Charge Estimation based on Domain-Transformation and Linear Discriminant Analysis
Carbone P.;De Angelis A.;Buchicchio E.;Santoni F.;Moschitta A.
2023
Abstract
This paper considers the estimation of the state-of-charge of rechargeable batteries based on a classifier trained using two methods. One method uses the values of the parameters in an equivalent circuit model, identified using a frequency-domain approach. The other method is based on a mathematical approximation of the battery voltage time-response to a given 3 s current signal. Classification resorts to a linear discriminant analysis classifier trained both by experimental data and by data obtained through augmentation methods. It is shown that the time-domain based classifier may achieve better performance in terms of probability of correct state-of-charge classification, using experiments of significant less duration than those associated with the usage of the frequency-domain experiments.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.