Battery cell voltage is an important evaluation index for electric vehicle condition estimation and one of the main monitoring parameters of the battery management system, and accurate voltage prediction is crucial for electric vehicle battery failure warning. Therefore, this paper proposes a novel hybrid gated recurrent unit and long short-term memory (GRU-LSTM) neural network to predict electric vehicle lithium-ion battery cell voltage. Firstly, Pearson coefficient correlation analysis is carried out to determine the input parameters of the neural network by analyzing the influence factors of the voltage parameters, and the hyperparameters of the neural network are determined through cross-validation to construct the lithium-ion battery single-unit voltage prediction model based on GRU-LSTM. Secondly, the voltage prediction accuracy and robustness of the GRU-LSTM model are verified by training the historical data of real vehicles in spring, summer, fall, and winter, combined with four different error indicators. Finally, the feasibility of the proposed method is verified by designing hierarchical warning rules based on the prediction data to realize the accurate warning of multiple voltage anomalies.

In order to achieve low carbon emissions in the power grid, the impact of new energy grid connections on the power grid should be reduced, as well as the peak-to-valley load difference caused by large-scale electric vehicle grid connections. This paper proposes a two-tier, low-carbon optimal dispatch master–slave game model involving virtual power plant operators as well as electric vehicle operators. Firstly, the carbon flow is tracked based on the proportional sharing principle, and the carbon emission factor during the charging and discharging process of electric vehicles is calculated. Secondly, the node carbon potential and time-sharing tariff are used to guide and change the charging behaviour of electric vehicles and to construct a master–slave game model for low-carbon optimal scheduling with the participation of multiple subjects, with economic scheduling at the upper level of the model and demand response scheduling at the lower level. Finally, the IEEE30 node system is used as an example to verify that the method adopted in this paper can effectively reduce the peak-to-valley difference of loads, reduce the carbon emissions of the grid, and reduce the cost of each participating entity.

The problem of electricity pricing for charging stations is a multi-objective mixed integer nonlinear programming. Existing algorithms have low efficiency in solving this problem. In this paper, a convex optimization algorithm is proposed to get the optimal solution quickly. Firstly, the model is transformed into a convex optimization problem by second-order conic relaxation and Karush–Kuhn–Tucker optimality conditions. Secondly, a polyhedral approximation method is applied to construct a mixed integer linear programming, which can be solved quickly by branch and bound method. Finally, the model is solved many times to obtain the Pareto front according to the scalarization basic theorem. Based on an IEEE 33-bus distribution network model, simulation results show that the proposed algorithm can obtain an exact global optimal solution quickly compared with the heuristic method.

The remaining useful life (RUL) prediction of lithium-ion batteries remains challenging for hybrid models due to high computational redundancy and hyperparameter sensitivity under complex operating conditions. To address these issues, this paper proposes a novel hybrid framework that integrates a CNN–BiLSTM–Attention network with a Crested Porcupine Optimizer (CPO). The key innovation lies in the simultaneous co-optimization of model structure and parameters: a group-level structured pruning strategy eliminates redundant convolutional kernels to reduce complexity, while the CPO algorithm dynamically optimizes critical hyperparameters (e.g., BiLSTM hidden nodes, attention dimensions) using RMSE as the fitness function. This dual optimization achieves a balance between model lightweighting and predictive accuracy. Experimental results on real-world electric vehicle datasets demonstrate that the optimized model reduces Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) by 60.7% and 55.2%, respectively, compared to the baseline CNN–BiLSTM–Attention model. Furthermore, the model maintains high robustness across multiple datasets (R² > 0.996), validating its strong generalization capability in small-sample and high-noise scenarios.