State-of-Charge assessment of Li-ion battery using Genetic Algorithm-Neural Network (GANN)
E. Cardelli, Fabio Crescimbini, Francesco Riganti Fulginei, Michele Quercio, Lorenzo Sabino
Abstract
The accurate estimation of State-of-Charge (SoC) is crucial for optimal performance and safe operation of lithium batteries. Traditional methods for SoC estimation have limitations in terms of robustness and accuracy, leading to the exploration of alternative techniques such as neural networks (NN). Neural networks are highly effective mathematical models that take inspiration from the organization and operation of the human brain, and their ability to handle complex nonlinear relationships makes them ideal for SoC estimation. The aim of this work is to train a NN with an optimized architecture for SoC predicting. In particular a Genetic Algorithm Neural Network (GANN) was used with three hidden layers to evaluate the state of charge of the lithium battery. The results show that an average error of 2% is riched on the test set. So the GANN method can be considered promising for this kind of evaluation.