Internal Resistance Estimation of Li-ion Batteries using Wavelet Analysis

A Battery Digital Twin From Laboratory Data Using Wavelet Analysis and Neural Networks

“…Different field applications require different aspects of battery modeling. Several models have been developed in the literature, depending on the targeted level of abstraction [8]: empirical models, which use functions to describe the dynamics of the battery [9]; electrochemical models, which offer insight to the internal chemistry of the battery and use complex partial differential equations to describe the dynamics of the battery [10], [11], [12], [13]; equivalent circuit models (ECMs), which model the dynamics of the battery with circuit components, such as voltage generator, resistors (R), and capacitors (C) [14], [15], [16]. ECM models are computationally lighter with respect to other models.…”

“…After several trials on different partitions of RC elements, it turns out that the best generalization results can be obtained by training an NN for each function R 0 , R 1 , C 1 , R 2 , C 2 , R 3 , C 3 , OCV. In this way, each NN needs only a few neurons (10)(11)(12)(13)(14)(15)(16)(17)(18)(19)(20) and the optimization can be carried out without interference from the other parameters. Other choices lead to bulk NNs with many neurons and with poor generalization.…”