Accurate prediction of remaining useful life (RUL) plays an important role in maintaining the safe and stable operation of Lithium-ion battery management systems. Aiming at the problem of poor prediction stability of a single model, this paper combines the advantages of data-driven and model-based methods and proposes a RUL prediction method combining convolutional neural network (CNN), bi-directional long and short-term memory neural network (Bi-LSTM), SE attention mechanism (AM) and adaptive unscented Kalman filter (AUKF). First, three types of indirect features that are highly correlated with RUL decay are selected as inputs to the model to improve the accuracy of RUL prediction. Second, a CNN-BLSTM-AM network is used to further extract, select and fuse the indirect features to form predictive measurements of the identified degradation metrics. In addition, we introduce the AUKF model to increase the uncertainty representation of the RUL prediction. Finally, the method is validated on the NASA dataset and the CALCE dataset and compared with other methods. The experimental results show that the method is able to achieve an accurate estimation of RUL, a minimum RMSE of up to 0.0030, and a minimum MAE of up to 0.0024, which has high estimation accuracy and robustness.

With its technical advantages of high speed, low latency, and broad connectivity, fifth-generation mobile communication technology has brought about unprecedented development in numerous vertical application scenarios. However, the high energy consumption and expansion difficulties of 5G infrastructure have become the main obstacles restricting its widespread application. Therefore, aiming to optimize the energy utilization efficiency of 5G base stations, a novel distributed photovoltaic 5G base station DC microgrid structure and an energy management strategy based on the Curve Fitting–Perturb and Observe–Incremental Conductance (CF-P&O-INC) Maximum Power Point Tracking (MPPT) algorithm from the perspectives of energy and information flows are proposed. Simulation results show that the proposed MPPT algorithm can increase the efficiency to 99.95% and 99.82% under uniform irradiation and partial shading, respectively. Under the proposed strategy, when the base station load changes drastically, the voltage fluctuation of the DC bus is less than 1.875%, and returns to a steady state within 0.07s, alleviating the high energy consumption of 5G base stations effectively and achieving coordinated optimization management of various types of energy in multi-source power supply systems.

The fourth industrial revolution, driven by the Energy Internet (EI), is having a profound impact on economic development and way of life. With the growth of EI networks, the integration of numerous energy devices poses challenges across different domains. To address this, we propose a self-adaptive NSGA-III algorithm (SA-NSGA-III) for multi-objective optimization of the EI topology, accounting for connectivity, robustness, and operational efficiency. We construct an initial scale-free topology based on real-world EI characteristics and optimize it while preserving its scale-free nature. The method incorporates an adaptive dynamic reference point generation strategy and an adaptive population selection mechanism. Experimental results demonstrate that SA-NSGA-III achieves a 29.5% fitness improvement, outperforming other multi-objective optimization algorithms in both optimization performance and convergence efficiency across various network scales and densities.