Accurate prediction of wind power ramp events is critical for ensuring the reliability and stability of power systems. However, most existing research on wind ramp events forecasting exhibits certain limitations, including: 1) extreme weather is marked by high meteorological variability, leading to unstable wind power output and frequent ramp events. However, its low occurrence frequency limits the availability of representative samples, while numerical weather prediction (NWP) suffers from large errors under such conditions. These factors together make ramp event prediction particularly challenging; 2) spatiotemporal correlation has not been fully considered, which restricts the improvement of forecasting accuracy. To address these limitations, this study introduces a day-ahead wind ramp events forecasting method considering extreme weather and spatiotemporal correlation information. First, cold wave events are precisely identified and the corresponding weather scenario dataset is expanded to enhance sample diversity and improve model robustness. Subsequently, a diffusion model is employed to correct NWP data, thereby providing more accurate input features for subsequent forecasting. Building upon this, the model further integrates spatiotemporal correlations among different power stations with local meteorological features of the target station, enabling refined extraction of predictive information for wind power ramp events. Finally, separate forecasting models are constructed for cold wave and conventional weather conditions to enhance adaptability and prediction accuracy. Validation on a real-world dataset in China confirms that our method outperforms the best existing benchmark by 2.7% in accuracy and 3.6% in F1-score.

To address the issues caused by a single evaluation index for ramping demand assessment in new power systems and the unclear evolution patterns of new energy penetration rates, this paper constructs a comprehensive assessment index system. Based on the data from a certain province’s power grid in 2023, different scenarios of new energy penetration rates and the proportions of wind and solar energy are analyzed. The results indicate that an increase in new energy penetration significantly raises the system’s ramping demand, with a greater impact on downward ramping; high proportions of photovoltaics lead to high ramping demand, while an increase in wind power can reduce ramping demand, especially in terms of a significant reduction in upward ramping demand. This research provides a reference for evaluating downward ramping demand and flexible resource allocation under high proportions of new energy integration.

Under China’s dual carbon goals, large-scale renewables integration challenges grid stability. With wind capacity reaching 440 million kW by 2024, static constraints prove inadequate. This study proposes a dynamic-constrained dispatching strategy using "limit calculation – multi-source dispatching – optimization". A Gaussian kernel-Copula-FCM based renewable model minimizes costs, yielding 14.7% cost reduction, 54.2% lower emissions, reduced curtailment, and better peak-shaving, with high scalability.

To address the complexity of energy storage evaluation and dispatch in ring-shaped heating networks within integrated electrical-thermal energy systems, this paper proposes a dynamic energy storage automation evaluation and optimal control method based on the heat current method. Leveraging graph theory, the ring-shaped network is decomposed into modular heat current models under supply and return water conditions. By combining time-domain simulation with a bisection iteration strategy, the passive energy storage capacity of the network is quantified in real-time. Using dynamic thermal storage characteristics as boundary conditions, a model predictive control (MPC) rolling optimization framework is constructed to generate day-ahead dispatch plans that balance thermal system requirements and wind power accommodation. The proposed method significantly reduces carbon emissions in integrated energy systems, enhances multi-energy synergy flexibility, and provides theoretical and technical pathways for integrating high-penetration renewable energy.

The proportion of new energy in the new power system is increasing year by year, and its uncertainty and volatility pose challenges to the operation of the new power system. Flexible resource regulation not only ensures the safe and stable operation of the system, but also serves as an important pillar in response to the green and low-carbon transformation of energy. Based on this situation, this article studies the coordinated and optimized operation of flexible resources in the new power system, analyzes the regulation of flexible resources on the power source side, load side, and energy storage side, establishes a quantitative model for flexibility demand, and takes the minimum total cost of flexible resources as the objective function. The hierarchical dual deep Q network algorithm is used to simulate and solve the problem

In high-latitude regions, the frequent occurrences of snowfall and snow coverage during winter can lead to significant fluctuations in photovoltaic power output, thereby triggering large-magnitude and long-duration ramp events, which pose risks to power grid dispatching. To address the insufficient consideration of snow conditions in existing studies, this paper proposes a two-stage indirect forecasting framework of "explicit snow condition modeling – power forecasting – event warning." First, a "snowfall–snow coverage mask" is constructed: the mask is set to 1 during snowfall periods and then linearly decays after snowfall to generate a continuous snow coverage score, characterizing the degree of residual snow obstruction. Then, the mask, along with meteorological features selected by mRMR feature selection considering Granger causality and historical power data, is input into a conditional attention Transformer model to complete day-ahead power forecasting. Subsequently, three ramp indicators are designed (15-minute slope R1, 4-hour maximum ramp rate R2, and 4-hour net variation D3). Within a recent 30-day sliding window, the quantile thresholds of the three ramp indicators are adaptively estimated. The binarized 0/1 sequences of these indicators are then combined using inverse-weighted soft voting, and continuous "1" segments are merged to form the predicted ramp events, resulting in the final ramp warning outcomes. Finally, predicted and actual events are matched using temporal intersection-over-union, and performance is evaluated. Validation based on real photovoltaic power plant data shows that the proposed method significantly improves power forecasting accuracy and ramp event warning performance under snowfall and snow coverage scenarios, verifying the effectiveness of the proposed approach.

Accurate regional distributed photovoltaics (PV) power probabilistic forecasting can provide comprehensive information support for the operation of active distribution networks. The mining and utilization of distributed PV spatiotemporal correlation information is an effective way for the improvement of power forecasting accuracy. However, several issues have been encountered in spatiotemporal correlation information mining for distributed PV forecasting: 1) traditional clustering studies are unable to reveal deep similarities between PV stations; 2) the power data contains a lot of redundant information, yet current deep learning feature extraction method does not adequately extract the similarity of implicit features and trend changes in time-series curves; 3) existing studies either fail to capture rich spatio-temporal correlation information during the modeling process, or ignore the dynamic changes of spatiotemporal correlation influenced by weather conditions. For this reason, this paper proposes a regional distributed PV power ultra-short-term probabilistic forecasting method based on dynamic hierarchical correlation modeling. First, a clustering method based on a deep similarity method is proposed to support targeted modeling of intra-region spatiotemporal correlations, which better extracts the intrinsic similarity of different PVs by deep learning model each distributed PV; next, feature extraction optimization on historical power data is performed by contrastive learning model to achieve effective extraction of the implicit features and trend changes, providing higher quality input features for subsequent forecasting models. Then, dynamic characterization of spatiotemporal correlations of distributed PVs is carried out by taking into account the effect of short-term changes in weather; on this basis, the hierarchical graph structure is constructed to model both the strong correlation within intra-region and the inter-subgraph correlation. Finally, the probabilistic forecasting model based on the Dynamic Hierarchical Graph Convolutional Neural (DHGCN) network and Mean Interval Score (MIS) regression is proposed, which mines the in-depth spatiotemporal correlation among PV power stations to enhance the accuracy of the forecasting. The effectiveness of the method is verified using an actual distributed PV power dataset.

With the increasing frequency of extreme climate events, power systems require effective post-disaster recovery strategies to enhance resilience. This paper proposes a generator start-up strategy that explicitly incorporates dynamic frequency security constraints. Key constraints, including the Rate of Change of Frequency and frequency nadir, are formulated to ensure system stability. To mitigate the modeling errors introduced by linearization, a compensation mechanism is developed to improve accuracy. Additionally, to avoid the computational complexity of dynamic line capacity constraints, these are excluded from the main optimization process and instead verified post-solution. If violations are detected, steady-state line power limits are dynamically adjusted to reserve sufficient margin for transient power. The objective function minimizes the costs of conventional generation and load shedding, subject to both secure operation constraints and dynamic frequency constraints. A case study on a modified IEEE 57-bus system demonstrates that the proposed strategy enables rapid load restoration and cost-efficient recovery, significantly improving system resilience and stability in the aftermath of disasters.

Under the background of “dual-carbon”, expanding renewable energy grid integration exacerbates grid net load volatility, and system climbing requirements escalate. In this paper, the problem of uncertain ramping demand prediction caused by net load prediction error in new power systems is investigated. First, the total system ramping demand calculation model is constructed, and the effects of deterministic and uncertain ramping demand on the total system ramping demand are analyzed. Secondly, a prediction model based on a CNN-LSTM hybrid neural network is proposed for the uncertain ramp-up demand, which extracts the spatial correlation features of the multi-source influencing factors through the convolutional layer, captures the dynamic evolution law in the time series by using the LSTM layer, and realizes the high-precision point prediction and reliable interval prediction by combining the quantile regression method. Finally, the actual operation data and forecast data of a provincial power grid are used for example verification, and the results show that the proposed model outperformed traditional models (SVM, RF, BPNN) and single deep learning models (CNN, LSTM) in point prediction performance, achieving higher prediction accuracy and validating the effectiveness of the spatio-temporal feature extraction module. In terms of interval prediction quality, compared with the histogram and QRF benchmark models, the proposed model achieves a significant reduction in the average width of the prediction interval, average upward ramp-up demand, and average downward ramp-down demand while maintaining 100% interval coverage. This demand realizes a better balance between prediction economic efficiency and safety, providing more reliable technical support for the precise assessment of uncertain ramp-up demand in new power systems.

Proton exchange membrane fuel cells (PEMFCs), recognized as promising sources of waste heat for space heating, domestic hot water supply, and industrial thermal applications, have garnered substantial interest owing to their environmentally benign operation and high energy conversion efficiency. Since the uniformity of oxygen diffusion toward catalytic layers critically governs electrochemical performance, this study establishes a three-dimensional, non-isothermal computational fluid dynamics (CFD) model to systematically optimize the cathode flow channel width distribution, targeting the maximization of power output through enhanced reactant homogeneity. Numerical results reveal that non-uniform flow channel geometries markedly improve oxygen distribution uniformity, reducing the flow inhomogeneity coefficient by 6.6% while elevating maximum power density and limiting current density by 9.1% and 7.8%, respectively, compared to conventional equal-width designs. There were improvements attributed to the establishment of longitudinal oxygen concentration gradients and we alleviated mass transfer limitations. Synergistic integration with gas diffusion layer (GDL) gradient porosity optimization further amplifies performance, yielding a 12.4% enhancement in maximum power density and a 10.4% increase in limiting current density. These findings validate the algorithm’s efficacy in resolving coupled transport constraints and underscore the necessity of multi-component optimization for advancing PEMFC design.

In response to the issue that the current power system has insufficient coordination ability of flexible resources and cannot meet the system's low-carbon economic operation process, this article deeply studies the low-carbon economic dispatch method of power system coordinated by composite energy storage. Firstly, the operation architecture of the hybrid storage system of energy and its operating characteristics and flexibility advantages are analyzed. Then, a low carbon economic dispatch power system's model with composite energy storage is given. The near-end strategy optimization algorithm is used to solve the low carbon economic dispatch model. Finally, the comparative analysis of different scenarios is set up to prove that the proposed low carbon economic dispatch method for a power system integrated with composite energy storage can boost the economy in the operation of the system and decrease carbon emissions.

The escalating penetration of renewable energy (RE) sources, notably photovoltaic (PV) and wind power, introduces substantial operational complexities for day-ahead scheduling and reserve management in modern power systems, owing to the inherent unpredictability of $\mathbf{R E}$ generation. To address the above challenges, this study proposes a collaborative optimization strategy for day-ahead scheduling and reserve configuration that explicitly accounts for RE uncertainty. The methodology is structured into three sequential stages: first, the probability distribution of RE forecast errors is derived through statistical analysis of historical data, providing a quantitative basis for characterizing uncertainty; second, reserve requirements are systematically determined based on the obtained error distribution to mitigate potential imbalances between supply and demand; and finally, a comprehensive cost-minimization model is developed, incorporating risk costs to hedge against uncertainties. The proposed model rigorously enforces constraints related to power balance, operational limits of thermal units, energy storage systems (ESSs), and grid interactions, while integrating reserve constraints to ensure system stability. Results of the simulations indicate that the proposed day-ahead scheduling framework enables an optimal allocation of reserve capacity, effectively balancing operational security and economic efficiency under uncertain RE conditions.

In the context of climate change and energy transition, the growing frequency of extreme weather events threatens the safety and stability of power systems. Given the limitations of existing research on load characteristic analysis and load forecasting during extreme weather events, this paper proposes a load-integrated forecasting model that accounts for extreme weather. First, an improved power load clustering method is proposed, combining Kernel PCA for nonlinear dimensionality reduction and an enhanced k-means algorithm, enabling both qualitative analysis and quantitative representation of load characteristics under extreme weather. Second, an optimal combination forecasting model is developed, integrating improved SVM and enhanced LSTM networks. Building upon the improved power load clustering algorithm, a load-integrated forecasting model considering extreme weather is established. Finally, based on the proposed load-integrated forecasting model, a time-series production simulation model considering extreme weather is constructed to quantitatively analyze the power and electricity balance risks of the system. Case studies demonstrate that the proposed integrated forecasting model can effectively analyze load characteristics under extreme weather and achieve more accurate load forecasting, which can provide guidance for the planning and operation of new power systems under extreme weather conditions.

With the increasing penetration of renewable energy generation in energy systems, power and district heating systems (PHSs) continue to encounter challenges with wind and solar curtailment during scheduling. Further integration of renewable energy generation can be achieved by exploring the flexibility of existing systems. Therefore, this study systematically explores the deep transfer modifications of a specific thermal power plant based in Liaoning, China, and the operational characteristics of the heating supply system of a particular heating company. In addition, the overall PHS operational performance is analyzed. The results indicate that both absorption heat pumps and solid-state electric thermal storage technologies effectively improve system load regulation capabilities. The temperature decrease in the water medium in the primary network was proportional to the pipeline distance. When the pipeline lengths were 1175 and 14,665 m, the temperature decreased by 0.66 and 3.48 °C, respectively. The heat exchanger effectiveness and logarithmic mean temperature difference (LMTD) were positively correlated with the outdoor temperature. When the outdoor temperature dropped to −18 °C, the heat exchanger efficiency decreased to 60%, and the LMTD decreased to 17.5 °C. The study findings provide practical data analysis support to address the balance between power supply and heating demand.

Accurate prediction of wind power ramp events is crucial for ensuring secure dispatch and efficient operation of power systems. However, existing studies suffer from three main limitations: first, the issue of small sample learning caused by scarce data under extreme weather conditions has not been adequately addressed; second, numerical weather prediction (NWP) exhibits significantly amplified errors under such extreme conditions; third, effective utilization of spatiotemporal correlations is lacking in current modeling approaches. To tackle these challenges, this paper proposes a wind power ramp event forecasting method that integrates an extreme weather adaptation mechanism with spatiotemporal correlation modeling. Specifically, a Generative Adversarial Network (GAN) is employed to augment samples in extreme weather scenarios, thereby enhancing model generalization under small-sample conditions. Subsequently, a Long Short-Term Memory (LSTM) network is introduced to correct errors in key NWP variables, improving the reliability of input features. Furthermore, a Graph Convolutional Network (GCN) is utilized to capture spatial dependencies among multiple wind farms, which is combined with temporal characteristics to construct a spatiotemporal joint prediction model. Test results demonstrate that the proposed method outperforms existing benchmark models in both ramp event identification and prediction accuracy.

In order to improve the frequency regulation performance of provincial power grid, considering the security and stability of provincial power grid under the CPS assessment standard and based on the characteristics of multi-stage participation of renewable energy in regulation, an evaluation method of CPS load frequency regulation capacity of provincial power grid with renewable energy was proposed. Firstly, the operation characteristics of the provincial grid system are analyzed, and the frequency regulation models of photovoltaic, biomass, diesel, energy storage and wind power are established. Then, the power deviation control of provincial power grid under CPS standard is studied. Finally, a fuzzy logic based method for evaluating CPS frequency regulation capability of provincial power grids is proposed, and an improved sine and cosine optimization algorithm is used to solve the evaluation model to evaluate the frequency modulation capability of provincial power grids considering the CPS assessment criteria.

The rapid development of renewable energy has brought us closer to achieving carbon peak and carbon neutrality goals, but it has also exacerbated the challenge of load fluctuation in power grids. To enhance grid stability and security, it is essential to implement effective peak shaving measures. Relying solely on various types of power generation to complement each other has limitations. In this paper, we propose a model and method for configuring the capacity requirements of large-scale liquid flow batteries involved in grid peak shaving, within the context of increasing demand for renewable energy integration. Our model utilizes an optimization approach with the objective of minimizing the operational costs of the peak shaving system. The model not only considers the technical constraint index of the system, but also innovatively introduces the reliability constraint of the probability of insufficient peak regulation and the reliability constraint of the mean value of insufficient peak regulation, and forms the optimal configuration model of peak regulation capacity of large-capacity flow battery assisted power system. The model takes the 24-hour load time series and wind power time series as inputs, employs the YALMIP toolbox, and utilizes the CPLEX solver to optimize the capacity and power requirements for large-scale liquid flow batteries assisting in renewable energy dispatch. A simulation analysis is conducted using a case study with a 24-hour load time series and wind power time series as inputs to evaluate the proposed model.

The heat storage evaluation of the heating network in the integrated power and thermal system plays a crucial role in improving the flexibility of the system. However, there is still a lack of energy storage evaluation methods for complex heating networks. Therefore, this paper proposes an automatic evaluation and utilization technology for heating network heat storage in integrated power and thermal systems, considering the passive heat storage characteristics and regulation capabilities of thermal systems, based on the heat current method. This technology automates and standardizes the modeling of branched heating networks in integrated power and thermal systems. By considering the characteristics of electrical and thermal transport processes, this automatic evaluation and control technology can maximize the heat storage potential and regulation capabilities of thermal systems.

With the continuous improvement of renewable energy penetration rate and the increase in heterogeneous energy demand, the safe and economic operation of multi-energy microgrids is becoming more and more important. The purpose of this paper is to study the optimal control strategy of multi-functional distributed energy storage system, draw on the concept of the Internet, build energy infrastructure from the bottom up, increase the flexible access and local consumption of distributed renewable energy through the open peer-to-peer interconnection and information energy fusion of similar energy autonomous units such as microgrids, realize its efficient and safe operation, reasonably evaluate the function and goals of energy storage in the distribution network, and configure and regulate the energy system through multi-objective optimization methods.