With the continuously increasing proportion of renewable energy integration, the structure of power grid networks has become increasingly complex. Under extreme weather conditions such as typhoons and hail, faults like line breaks or information disruptions can occur in the power grid, imposing significant burdens and risks on the economic and reliable operation of the power system. However, existing methods still focus on the allocation of human repair teams, with insufficient utilization of flexible resources within the system, resulting in low efficiency in restoring power supply to the power system. To address this challenge, this paper proposes a resilience enhancement strategy for the power system under typhoon scenarios. It leverages active resources on the grid side and fully exploits the flexibility of both the supply and demand sides to enhance the resilience of the power system. Firstly, this paper aims at the economic operation of the power system, taking into account the physical and operational constraints of both the supply and demand sides, including power flow constraints, mobile energy storage system (MESS) transfer constraints, and phase-shifting transformer (PST) regulation constraints. Meanwhile, an improved grasshopper optimization algorithm is introduced to achieve efficient and rapid problem-solving. Finally, the effectiveness and feasibility of the proposed method are demonstrated through validation using an improved IEEE-33 bus test system. Through analysis, the total system load loss was reduced by 75.6%, with the maximum load loss during the typhoon decreasing by 72.4%. The approach enables real-time response to the dynamic impacts of typhoons, swiftly stabilizes load fluctuations, and significantly enhances the resilience of the power system.

Accurate and efficient short-term load forecasting is crucial for the secure and stable operation and scheduling of power grids. Addressing the inability of traditional Transformer-based prediction models to capture symmetric correlations between different feature sequences and their susceptibility to multi-scale feature influences, this paper proposes a short-term power distribution network load forecasting model based on an enhanced Komodo Mlipir Algorithm (KMA)—Multivariate Variational Mode Decomposition (MVMD)-Crossformer. Initially, the KMA is enhanced with chaotic mapping and temporal variation inertia weighting, which strengthens the symmetric exploration of the solution space. This enhanced KMA is integrated into the parameter optimization of the MVMD algorithm, facilitating the decomposition of distribution network load sequences into multiple Intrinsic Mode Function (IMF) components with symmetric periodic characteristics across different time scales. Subsequently, the Multi-variable Rapid Maximum Information Coefficient (MVRapidMIC) algorithm is employed to extract features with strong symmetric correlations to the load from weather and date characteristics, reducing redundancy while preserving key symmetric associations. Finally, a power distribution network short-term load forecasting model based on the Crossformer is constructed. Through the symmetric Dimension Segmentation (DSW) embedding layer and the Two-Stage Attention (TSA) mechanism layer with bidirectional symmetric correlation capture, the model effectively captures symmetric dependencies between different feature sequences, leading to the final load prediction outcome. Experimental results on the real power distribution network dataset show that: the Root Mean Square Error (RMSE) of the proposed model is as low as 14.7597 MW, the Mean Absolute Error (MAE) is 13.9728 MW, the Mean Absolute Percentage Error (MAPE) reaches 4.89%, and the coefficient of determination (R²) is as high as 0.9942.

To address the challenges of climate change mitigation and operational flexibility in active distribution networks (ADNs) amid high renewable energy penetration, this paper proposes a low-carbon economic dispatch framework integrating demand-side carbon regulation and cyber–physical system (CPS)-enabled shared energy storage. First, a consumer-side emission penalty mechanism is developed by fusing a carbon emission flow (CEF) model with price elasticity coefficients. This mechanism embeds carbon costs into end-user electricity pricing, guiding users to adjust consumption patterns (e.g., reducing usage during high-carbon-intensity periods) and shifting partial carbon responsibility to the demand side. Second, a CPS-based shared energy storage mechanism is constructed, featuring a three-layer architecture (physical layer, control decision layer, security layer) that aggregates distributed energy storage (DES) resources into a unified, schedulable pool. A cooperative, game-based profit-sharing strategy using Shapley values is adopted to allocate benefits based on each DES participant’s marginal contribution, ensuring fairness and motivating resource pooling. Finally, a unified mixed-integer linear programming (MILP) optimization model is formulated for ADNs, co-optimizing locational marginal prices, DES state-of-charge trajectories, and demand curtailment to minimize operational costs and carbon emissions simultaneously. Simulations on a modified IEEE 33-bus system demonstrate that the proposed framework reduces carbon emissions by 4.5–4.7% and renewable energy curtailment by 71.1–71.3% compared to traditional dispatch methods, while lowering system operational costs by 6.6–6.8%. The results confirm its effectiveness in enhancing ADN’s low-carbon performance, renewable energy integration, and economic efficiency.

Safety-critical scenarios are infrequent in natural driving environments but hold significant importance for the training and testing of autonomous driving systems. The prevailing approach involves generating safety-critical scenarios automatically in simulation by introducing adversarial adjustments to natural environments. These adjustments are often tailored to specific tested systems, thereby disregarding their transferability across different systems. In this paper, we propose AdvDiffuser, an adversarial framework for generating safety-critical driving scenarios through guided diffusion. By incorporating a diffusion model to capture plausible collective behaviors of background vehicles and a lightweight guide model to effectively handle adversarial scenarios, AdvDiffuser facilitates transferability. Experimental results on the nuScenes dataset demonstrate that AdvDiffuser, trained on offline driving logs, can be applied to various tested systems with minimal warm-up episode data and outperform other existing methods in terms of realism, diversity, and adversarial performance.