Deep learning approach towards accurate state of charge estimation for lithium-ion batteries using self-supervised transformer model.

Accurate state of charge (SOC) estimation of lithium-ion (Li-ion) batteries is crucial in prolonging cell lifespan and ensuring its safe operation for electric vehicle applications. In this article, we propose the deep learning-based transformer model trained with self-supervised learning (SSL) for end-to-end SOC estimation without the requirements of feature engineering or adaptive filtering. We demonstrate that with the SSL framework, the proposed deep learning transformer model achieves the lowest root-mean-square-error (RMSE) of 0.90% and a mean-absolute-error (MAE) of 0.44% at constant ambient temperature, and RMSE of 1.19% and a MAE of 0.7% at varying ambient temperature. With SSL, the proposed model can be trained with as few as 5 epochs using only 20% of the total training data and still achieves less than 1.9% RMSE on the test data. Finally, we also demonstrate that the learning weights during the SSL training can be transferred to a new Li-ion cell with different chemistry and still achieve on-par performance compared to the models trained from scratch on the new cell.

Fully decentralized robust backstepping voltage control of photovoltaic systems for DC islanded microgrids based on disturbance observer method.

In this paper, a new backstepping-based nonlinear technique for control of photovoltaic systems in DC islanded microgrids (MGs) is proposed. In contrast to most existing droop/non-droop control strategies that require an exact model of the system including line impedances, loads, other distributed generation units (DGUs) parameters, and even the MG configuration, the proposed method is taking dynamics and uncertainties into account using a designed disturbance observer. Moreover, the proposed method rapidly reaches the reference values and exhibits a more accurate robust performance using local quantities measurement, irrespective of parametric uncertainties, unmodeled dynamics, unknown loads, disturbances, and the number/structure of DGs within the MG. Finally, a low-voltage DC MG is built where the robust performance of the proposed method for different operating conditions including load variation, tracking capability, nonlinear loads, and plug-play of DGs is verified.