Effective lithium-ion battery pack charging is of extreme importance for accelerating electric vehicle development. This paper derives an optimal charging control strategy with a leader-followers framework for battery packs. Specifically, an optimal average state-of-charge (SOC) trajectory based on cells' nominal model is first generated through a multi-objective optimization with consideration of both user demand and battery pack's energy loss. Then, a distributed charging strategy is proposed to make the cells' SOCs follow the pre-scheduled trajectory, which can effectively suppress the violation of the safety-related charging constraints through on-line battery model bias compensation. This work highlights the superiorities of the proposed leader-followers-based charging framework that combines the off-line scheduling and on-line closed-loop regulation for battery pack charging, which brings benefits to significantly reduce the computational burden for the charger controller as well as improve the robustness to suppress the negative impact caused by the cell's model bias. Extensive illustrative results demonstrate the effectiveness of the proposed optimal charging control strategy.

Effective cell equalization is of extreme importance to extract the maximum capacity of a battery pack. In this paper, two cell balancing objectives, including balancing time reduction and cells' temperature rise suppression, are taken into consideration simultaneously. Furthermore, hard constraints are imposed on the cells' state-of-charge levels, currents, and equalizing currents. Based on a developed module-based cell-to-cell balancing system model, a multi-objective constrained optimization problem is formulated, which aims at the coordinated control of all equalizers rather than individually controlling the equalizer for its two adjacent cells' equalization. Next, a hierarchical cell equalizing control approach is proposed, where the module-level controlled equalizing currents are first designed at the top layer and then the cell-level equalizers are controlled for each battery module in parallel at the bottom layer. The designed hierarchical structure significantly reduces the computational burden, making the cell equalizing algorithm more implementable in real-time. Following the Lyapunov stability analysis, the convergence of the designed cell equalizing control algorithm is proved. Illustrative results demonstrate that the balancing time can be reduced by up to 29.8% compared with the decentralized equalizing control.

Deep learning has become a powerful tool to automatically classify medical hyperspectral images (MedHSIs) for the diagnosis of various tumors such as cancer. These deep learning based classification approaches consist of both feature extraction and disease prediction, which are independent of each other. Therefore, the extracted features may be incompatible with the used classifier for prediction. To remedy such deficiency, in this work, we propose a novel deep method, i.e ., deep margin cosine autoencoder (DMCA), for the MedHSI classification, which provides solid support to diagnose tumors. To be specific, first , we consider deep autoencoder (DAE) network as the basic framework for the feature extraction of the MedHSI, and the soft-max classifier is introduced as the output layer to predict the results. Second , to report the features well compatible with the classifier, the importance of the soft-max is added as a constraint into the DAE network. Further, a cosine margin is introduced to enhance the discrimination of different feature clusters. In addition, we also design an optimization scheme to the solutions of the DMCA model. Third , a two-stage training strategy is presented to train the built DMCA network. After completing the training, unknown tissues ( i.e ., unlabeled samples) can be specified to determine whether they are tumors or not. Sufficient experimental results have been provided to validate that our DMCA method achieves better classification performance and higher tumor diagnosis accuracy compared with some advanced approaches.

The state-of-charge (SOC) is a fundamental indicator representing the remaining capacity of lithium-ion batteries, which plays an important role in the battery’s optimized operation. In this paper, the model-based SOC estimation strategy is studied for batteries. However, the battery’s model parameters need to be extracted through cumbersome prior experiments. To remedy such deficiency, a recursive least squares (RLS) algorithm is utilized for model parameter online identification, and an adaptive square-root unscented Kalman filter (SRUKF) is designed to estimate the battery’s SOC. As demonstrated in extensive experimental results, the designed adaptive SRUKF combined with RLS-based model identification is a promising SOC estimation approach. Compared with other commonly used Kalman filter-based methods, the proposed algorithm has higher precision in the SOC estimation.