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.

Low carbon operation of power systems is a key way to achieve the goal of energy power carbon peaking and carbon neutrality. In order to promote the low carbon transition of energy and power and the coordinated and optimized operation of distributed energy sources in virtual power plants (VPP), this paper proposes a framework for collaborative utilization of pumped storage–carbon capture–power-to-gas (P2G) technologies. It also constructs a multi-time scale low carbon economic dispatch model for VPP to minimize the internal resource operation cost of VPP in each time period. During the intraday scheduling stage, the day-ahead scheduling results as the planned output and the energy flow is then dynamically corrected at a short-term resolution in the framework. This allows for the exploration of the low-carbon potential of each aggregation unit within the virtual power plant. The results of the simulation indicate that the strategy and model proposed in this paper can effectively encourage the consumption of renewable energy sources, promote the low-carbon operation of power system power, and serve as a valuable reference for the low-carbon economic operation of the power system.

Accurate estimation of the state of charge (SOC) of lithium-ion batteries is critical for assessing the safety and remaining range of electric vehicles. However, due to the complex and variable operating environment of batteries and their highly nonlinear internal mechanisms, achieving high-precision SOC prediction remains a central challenge in current research. To this end, this paper proposes a nonlinear Hammerstein model based on the Hippopotamus Optimization Algorithm (HO) to optimize the backpropagation neural network, thereby enhancing the accuracy of SOC prediction. The HO-BP-Hammerstein model optimizes the BP neural network architecture using the Hippopotamus Algorithm and conducts SOC prediction accuracy tests on real-world data. Experimental results demonstrate the superiority of the proposed method through comparative accuracy analysis of various SOC prediction approaches under different operating conditions, confirming its significant engineering application value.

This paper proposes a Stackelberg game-based collaborative optimization strategy for Park-Level Integrated Energy Systems (PIESs) operating in carbon and green certificate markets. The strategy addresses interest conflicts and low-carbon transition challenges in multi-agent optimization by integrating a carbon capture, power-to-gas, and liquid carbon dioxide energy storage technology chain. Innovatively integrates LCES into the CCS-P2G-LCES chain, achieving internal carbon cycling and energy storage. First, a market environment for PIESs integrating carbon trading and green certificate trading is constructed, and a deeply coupled low-carbon technology chain model of CCS-P2G-LCES is established to realize internal carbon resource cycling and energy time shifting. Second, a one-leader, multiple-follower Stackelberg game framework is developed with the Integrated Energy Service Provider (IESP) as the leader and the User Load Aggregator (ULA) and Electric Vehicle Aggregator (EVA) as followers. The IESP guides demand response on the user and electric vehicle sides by formulating differentiated energy prices. On this basis, a collaborative optimization dispatch model is constructed with the objective of maximizing the comprehensive revenue of the IESP. Finally, case study analysis verifies that the proposed method not only enhances operational revenue and reduces user energy costs, but also significantly reduces system carbon emissions and improves renewable energy consumption rates. The results demonstrate the feasibility and superiority of integrating market mechanisms, low-carbon technologies, and multi-agent game-based collaborative optimization.

In view of the carbon emission reduction and new energy consumption problems in the integrated energy system (IES), this paper, for the first time, combines power to gas (P2G) with liquid carbon dioxide energy storage (LCES) and takes demand response (DR) into account simultaneously to construct a new type of IES capacity configuration optimization model. Firstly, based on the operation characteristics and coupling features of various devices within the system, the IES model was constructed. Meanwhile, the electricity, heat, and cold DR models were, respectively, established according to price-based and incentive-based methods. Finally, a two-layer collaborative optimization configuration model was built, with the upper layer aiming to minimize the annual total cost of the system and the lower layer aiming to minimize the annual operation cost of the system. Through case studies, the effectiveness of the established model was verified, and the impacts of DR, P2G, and LCES on the system capacity configuration results, economic efficiency, and environmental friendliness were investigated. Additionally, the impact of natural gas prices on the system optimization results was studied. The results showed that considering LCES and P2G could reduce the cost of the IES by 7.26% and the carbon emissions of the system by 31.03%, verifying the effectiveness of the proposed method and providing a feasible solution for IES capacity configuration.

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.