Against the backdrop of continuously increasing renewable energy penetration in the power distribution sector, the coexistence and energy interaction of multiple microgrids will form a microgrid cluster. Its operational optimization requires balancing economic efficiency, low-carbon performance, and multi-agent coordination. This paper focuses on grid-connected microgrid clusters, constructing a low-carbon economic dispatch model that incorporates time-of-use electricity prices, demand response, and carbon trading costs. Under the constraints of demand response, energy storage operation, microgrid power balance, and tie-line capacity limits, the model aims to minimize the total operational cost of the cluster through day-ahead optimal scheduling. By treating the interactive power of microgrids as coupling variables, a distributed solution method based on the alternating direction method of multipliers is proposed. Finally, an example analysis derives low-carbon economic operation strategies and compares the costs and carbon emissions of each microgrid under scenarios with and without distributed trading.

Peri-urban ammonia plants are often treated as rigid, carbon-intensive industrial loads. However, when retrofitted with renewable energy, they can serve as green flexibility resources that enhance the sustainability and resilience of urban power supply. Therefore, this study proposes a multi-objective, distributional robust planning framework that jointly optimizes flexible capacity allocation and economic configuration of a renewable power-to-ammonia (RePtA) system under uncertain scenarios. A Gaussian copula–based scenario generator preserves the joint dependence among wind generation, solar irradiance, and locational marginal prices, and a resource endowment index is used to identify renewable-deficit stress. To capture seasonal buffering effects without extending the time horizon to a full year, long-term hydrogen storage is modelled using cross-scenario coupling constraints on inventory dynamics. Results show that combining flexibility with long-horizon hydrogen buffering stabilizes ammonia production during stress regimes and provides 74.92 MW of flexible capacity for demand-side response. Under the representative deficit stress scenario, the coupled design reduces daily grid imports from 2.24 to 2.32 GWh to 1.39 GWh and cuts daily CO2 emissions from 1.23 to 1.28 kt to 0.56 kt, indicating that long-horizon buffering mitigates compound exposure to renewable scarcity. These findings suggest that integrating flexibility with long-duration storage coupling is a preferred pathway toward cost-effective, lower-carbon RePtA systems.

With the low-carbon transition of ports and the increasing penetration of renewable energy, port integrated energy stations face significant challenges in coordinating shore power supply, hydrogen refueling demand, and renewable energy accommodation under uncertainty. To address this issue, this paper proposes a coordinated scheduling framework considering multi-path green hydrogen utilization and renewable uncertainty. A coupled electricity-hydrogen-heat-gas-carbon integrated energy model is established by integrating PEM electrolysers, hydrogen storage systems, fuel cells, methanation reactors, combined heat and power units, and carbon capture facilities. Furthermore, a two-stage Wasserstein distributionally robust optimization (W-DRO) model is developed to handle uncertainties in renewable generation and port load demand using limited historical data. Comparative case studies demonstrate that the proposed strategy effectively improves renewable energy accommodation capability, reduces operating cost and carbon emissions, and enhances operational robustness under uncertain conditions. The results verify that coordinated green hydrogen utilization can significantly improve the flexibility and low-carbon performance of port integrated energy systems.

The transition toward decarbonized energy systems relies heavily on the large-scale integration of wind power, introducing stochastic volatility that challenges traditional grid planning paradigms. While standard stochastic optimization relies on scenario generation, capturing extreme scenarios—rare, high-impact events located in the distribution's long tail—remains a formidable challenge. Conventional Generative Adversarial Networks (GANs) often suffer from "distribution smoothing," failing to reproduce the sharp gradients and complex morphologies of extreme events. This paper presents a comprehensive methodology: a Conditional Wasserstein GAN with Gradient Penalty (cWGAN-GP) augmented by Persistent Homology (PH) constraints. By integrating Topological Data Analysis (TDA) to extract robust geometric invariants from wind power time series, and utilizing a differentiable proxy network (TopoNet), this approach enforces morphological fidelity in generated scenarios. The analysis demonstrates that this topologically regularized framework significantly outperforms baselines in reproducing the statistical, temporal, and physical characteristics of extreme wind power events, providing a critical tool for robust power system risk assessment.

The limited driving range of electric vehicle (EV) on long distance travel, such as highways, is a major barrier to their widespread adoption. The construction of highway charging stations has been an important development plan for governments. Analyzing and simulating the charging load on highways is essential for supporting the development of EV charging station and infrastructure. This paper examines the spatiotemporal distribution of highway traffic flow using a modified Cell Transmission Model, which is based on the fluid dynamics of traffic flow. Additionally, a charging load simulation method designed to handle excessive charging demand periods is introduced, utilizing the Stationary Backlog Carryover approach. This method effectively simulates the continuous backlog of EVs charging demand that frequently occurs during holidays and peak periods. The validity of the proposed method is demonstrated using data from a highway in Shandong, China, highlighting its potential to support EV infrastructure planning and pricing strategy.

High renewable energy penetration creates significant operational challenges for power systems, especially during extreme weather that disrupts supply-demand balance. This study introduces a framework that integrates extreme scenario identification, data augmentation, and power balance optimization. It defines extreme wind speed events, such as sudden drops, surges, and persistent anomalies, and uses a sliding-window algorithm to extract these events from historical meteorological data. To address the scarcity of extreme samples, a new data augmentation method combines the Wasserstein Generative Adversarial Network with Gradient Penalty (WGAN-GP) and iterative distribution shifting. This approach focuses the generated data on distribution tails while preserving diversity and temporal consistency. An optimization model, which includes various generation resources, energy storage, and load shedding, is developed to assess system flexibility under extreme conditions. Case studies on the projected 2030 Northeast China Power Grid show that the augmentation method expands extreme scenario datasets from 150 to 1000 samples, maintains extremity and temporal consistency, and reveals that wind curtailment rises sharply above 70% renewable share, with storage systems providing key flexibility in high-output scenarios.

Renewable power-to-ammonia (RePtA) plants face economic challenges due to wind-solar resource intermittency and deficit episodes. To address this, this study proposes a multi-objective stochastic programming framework that integrates hybrid scenario generation (normal and deficit scenarios), Conditional Value-at-Risk (CVaR), and supply reliability indicators. To make reliability interpretable from an engineering perspective, this paper defines and operationalizes a set of ammonia-specific supply indices that quantify the frequency, magnitude, and severity of production shortages. The model endogenously incorporates oxygen co-product revenue and carbon emission reduction benefits, and yields Pareto frontiers of economic-reliability trade-offs. A case study indicates that oxygen sales and carbon emission reduction materially improve economics and reduce tail risk, with revenue offsetting 12%-17% of the levelized cost of ammonia. Scenario analysis reveals pronounced reliability degradation under deficits: relative to normal scenarios, the expected duration of production loss rises from single-digit hours to 60 h and the unserved ammonia mass increases from 10 t to 200 t in severe deficit scenarios, while the instantaneous shortage ratio is capped at 46% by process constraints. The expected annual profit increases nearly linearly with the carbon price and reaches a break-even point at about 142 CNY/t-CO2. This work provides a reproducible methodological framework for risk-aware and reliable RePtA planning, quantifying how carbon policy, by-product monetization, and storage coordination jointly improve both economics and supply security.

The share of installed renewable capacity is expected to continue rising, driven by environmental and decarbonization goals. However, resource profiles and development conditions vary substantially across regions. This paper proposes a cooperative game framework for energy storage capacity planning to enhance cross-regional resource complementarity and utilization capabilities. First, the method combining K-Medoids and shapeDTW is employed to cluster planning scenarios based on the characteristics of renewable energy and load. Then, under the cooperative game framework, energy storage capacity planning and optimize operation method in cross-regional power system is developed to reduce energy storage planning and system operating costs across regions while preserving the privacy of each region. Finally, the cross-regional resource compensation strategy based on Nash bargaining is established to achieve a win-win outcome within the cooperative game framework. A decentralized distributed solution method using the enhanced Alternating Direction Method of Multipliers (ADMM) is adopted to solve the cross-regional energy storage capacity planning and resource compensation strategies. The case study utilizes the modified HRP 38-bus power system. Numerical results demonstrate that the proposed cooperative game framework improves the utilization of cross-regional renewable energy and reduces the planning and operating costs of energy storage in the cross-regional power system. Furthermore, the results confirm the cost advantages achieved by each region under the cooperative game framework.

With the rapid growth in electric vehicle (EV) adoption, highway charging stations (HCSs) composed of photovoltaic generation, energy storage systems, and charging piles have become an effective solution to alleviate range anxiety for long-distance EV travel and reduce carbon emissions in the transportation sector. This study investigates how highway microgrid operator (HMGO) can design distributed pricing strategies for multiple HCSs along highways to maximize their overall profit in the coupled electricity-carbon market. A Stackelberg-evolutionary joint game model is proposed. In the upper Stackelberg game, the interaction between the leader (HMGO) and followers (EVs) is modeled through electricity price and charging demand decisions. In the lower evolutionary game, the charging behaviors among EV users are modeled considering their bounded rationality in decision-making. The proposed approach quantifies the carbon reduction contributions of both HCSs and EVs into tradable carbon emission allowances, enabling their participation in the carbon market and guiding optimal charging price strategies. Case studies demonstrate that: (1) Compared with unified pricing and electricity-market-only pricing, the proposed pricing strategy increases the total profit of HMGO by 3.1% and 26.2%, respectively, while reducing the PV curtailment rate by 2.22% and 3.52%. (2) Carbon trading further enables the HMGO to lower the average EV charging price by 12.3%, thereby attracting additional charging demand and enhancing carbon trading profit. (3) The carbon trading price and the penetration rate of home charging piles significantly affect the optimal pricing and profitability, underscoring their importance in highway microgrid pricing and operation.

As the terminal responsible for delivering power to end users, the active distribution network (ADN) plays a critical role in modern power systems. However, renewable energy (RE) forecasting errors introduce inherent trade-offs among RE accommodation, power supply adequacy, and economic cost. Thus, this paper proposes a “predict-then-optimize” framework for ADN operation under uncertainty, in which forecasting residual information provides data-driven support for the distributionally robust optimization (DRO) model in downstream ADN decision-making. For RE and load forecasting, a large model for multivariate time series forecasting (LM-MTSF) is fine-tuned using meteorological and power-related features. The model exploits large-scale time-series pretraining to enable transferability and long-range dependency modeling. For optimization, the rolling forecast errors generated by LM-MTSF are used to construct the data-driven Wasserstein ambiguity set. Based on this set, a data-driven Wasserstein DRO (DWDRO) model is formulated for ADN scheduling with power-flow constraints and then reformulated through dual transformation into a tractable optimization problem. Case studies based on modified IEEE 33-bus and IEEE 69-bus ADNs demonstrate the applicability of the proposed “predict-then-optimize” framework, verify the effectiveness of fine-tuned LM-MTSF forecasts and rolling forecast error analysis in providing data-driven support for DRO. The results show that the proposed approach reduces unexpected power shortages and enhances overall economic performance.

The widespread adoption of electric vehicles (EVs) offers substantial environmental benefits but presents significant challenges for power system management due to the inherent uncertainties in EV users' behavior and charging demands. To address these issues, this study proposes a real-time nested scheduling model with an adaptive charging-discharging strategy for community, supermarket, and street parking facilities. The model employs an adaptive approach where EV users provide essential information upon arrival at the microgrid (MG), choosing between unordered or ordered participation based on their preferences. A priority-based two-stage schedule planning strategy is implemented, first planning the overall charging-discharging power for the EV cluster, then refining individual EV power based on priorities. For EVs requiring early departure, priorities are adjusted to ensure the expected State of Charge is reached before leaving. The model features a real-time nested framework with a bi-level optimization outer layer to coordinate conflicting interests among entities and an inner layer for EV power allocation. To address high uncertainties, a Model Predictive Control algorithm is employed, incorporating Karush-Kuhn-Tucker conditions, strong duality theorem, and linearization techniques. Simulation results demonstrate the model's superior operational efficiency, reliability, and robustness while achieving a win-win outcome for both the MG operator and EV cluster.

With the rapid progress of energy transition and the expansion of electricity markets, renewable energy operators are exposed to power fluctuations and electricity price uncertainties, which may lead to bidding deviations and revenue risks. The complementary operation of pumped storage with wind and photovoltaic generation can effectively improve renewable energy utilization and enhance profitability. This paper develops a day-ahead bidding optimization model for the WSPSJOE (Wind-Solar-Pumped Storage Joint Operation Entity). The model incorporates robust optimization to characterize generation uncertainty and employs CVaR (Conditional Value at Risk) to quantify price-related risks. Comparative analyses between independent and joint operation modes, as well as sensitivity studies on risk preference coefficients and robustness parameters, are conducted to validate the feasibility and effectiveness of the proposed model. The results demonstrate that accounting for both renewable generation and price uncertainties in the day-ahead bidding stage significantly enhances the operational robustness and revenue reliability of the WSPSJOE.

Solar-assisted carbon capture power plants (SACCPPs) leverage solar thermal energy to mitigate power output loss and expand the net output range of carbon capture units, enhancing energy economy, reserve capacity, and carbon emission reduction. However, assessing the flexible operation benefits of SACCPPs in power systems with high wind power penetration is challenging due to complex thermodynamic models and limited risk assessment methods. This study addresses these gaps by proposing an innovative modeling approach and benefits evaluation framework. First, a linear flexible operation model is developed, focusing on the technical features of SACCPPs relevant to power system operation and scheduling. This model elucidates the intercoordination in power generation, carbon capture, and thermal storage. The operating ranges of various carbon capture power plants are quantitatively analyzed using a two-dimensional coordinate diagram, highlighting the flexible regulation advantages of SACCPPs. Second, an interval-enhanced CVaR method is introduced, which considers random variables with unknown probability distributions, refining the current CVaR-based knowledge. This method is used for a quantified risk assessment to evaluate supply-demand imbalance risks in power systems, providing a foundation for assessing SACCPPs' risk mitigation benefits. Third, a risk-aware operation scheduling model is developed to explore SACCPPs' capability in enhancing the system's energy economic benefits, risk mitigation, and carbon emissions reduction. This model aids energy administration in evaluating system gains from SACCPPs and in developing rational system-wide planning and retrofit projects. Finally, numerical simulation and sensitivity analysis results on the modified IEEE-39 bus system validate the robust adaptability and effectiveness of the proposed models and methods.

With the large-scale integration of electric vehicle (EV) loads into the power grid, new challenges have emerged for power system operation. With the development of vehicle-to-grid (V2G) technology, the flexibility of EV resources has been significantly enhanced. Under the time-of-use (TOU) pricing mechanism, designing reasonable charging and discharging strategies for EVs can provide multiple benefits to the power system. How to achieve optimal scheduling at the system level in terms of economy and security, while meeting individual user demands, has become a key research focus. This paper proposes a two-stage optimal scheduling strategy for EV charging and discharging considering V2G. In the upper-level model, peak shaving and loss reduction are set as optimization objectives, and a dispatching mechanism based on typical charging/discharging profiles is established. An improved optimization algorithm, Levy flight-based Simulated Annealing Particle Swarm Optimization (LSAPSO), is introduced to enhance the performance in terms of global search ability and convergence efficiency. The lower-level model targets individual EV users and employs a linear programming approach to realize refined control of charging and discharging, taking into account electricity demand, battery life, and scheduling responsiveness. Simulation experiments based on the IEEE 33-bus distribution network system demonstrate that the proposed strategy effectively reduces load peak-valley differences, minimizes network loss, and improves economic efficiency and operational security.

Driven by the global carbon neutrality target, the low-carbon transformation of power system and the synergy of carbon markets have emerged as critical issues. This study focuses on the value reconstruction and cost-sharing optimization of thermal power in the coupled electricity-carbon multi-market environment. It quantitatively analyzes the evolution of thermal power revenue structures and carbon cost transmission paths by constructing a multi-case simulation model based on three cases: the baseline case, the carbon pricing case, and the synergistic optimization case.

The development of low-carbon power systems critically depends on incorporating renewable energy sources with emission mitigation technologies. Nevertheless, system reliability faces substantial challenges from the inherent variability in both renewable energy production and consumer load demand. This study addresses these challenges in coastal next-generation power systems by introducing CCS technology and a tiered carbon trading mechanism to reduce emissions and enhance renewable energy accommodation. A low-carbon economic dispatch model incorporating fuzzy chance constraints is proposed to handle uncertainties on both generation and load sides. Deterministic constraints are relaxed into fuzzy system constraints and processed using triangular fuzzy parameters. The model is solved using GUROBI. Case studies validate the model's effectiveness in balancing economic efficiency and system reliability under different confidence levels, providing a theoretical foundation to achieve carbon-efficient performance in next-generation coastal power grids.

In the face of energy crises and environmental deterioration, wind power has become an important solution to these problems. However, wind power has characteristics such as volatility, which poses difficulties for the scheduling of systems containing wind power. To address this, demand-side response, as a flexible resource that responds quickly and has high economic benefits, offers an alternative pathway for enhancing wind power utilization while reducing the system's operating costs. Accordingly, this study develops a day-ahead time-of-use electricity pricing PDR model grounded in the elasticity theory of electricity demand, as well as an IDR model considering capacity compensation and electricity compensation. Furthermore, this paper takes into account the costs of thermal power, IDR costs, carbon emission costs, load shedding costs, and related constraints, and establishes an optimization operation model for a system containing wind power considering demand-side response. Finally, through a 10-machine system example, a series of comparative analyses are conducted to illustrate the utility of different types of demand response for the operation of systems with wind power.

This paper presents a two-stage robust optimization frame-work for the economic scheduling of highway charging station (HCS) equipped mobile charging vehicles (MCV), considering uncertainties in PV generation and EV arrival rate. The proposed model integrates day-ahead MCV leasing decisions with intra-day operational scheduling. A master–subproblem decomposition approach is employed, where the bilinear term arising from uncertainty and decision variables is linearized, and the resulting problems are solved using a C&CG algorithm. Case studies demonstrate that under high-traffic conditions, MCV deployment significantly increases the charging capacity and total revenue of HCS. The robust optimization approach outperforms deterministic scheduling by providing stronger resilience against low PV generation and fluctuations in EV arrival rate, with only a marginal reduction in total revenue. These results highlight the effectiveness of MCV integration and robust scheduling for improving both the economic and operational reliability of HCSs under uncertainty.

An optimization model is proposed in this paper to allocate shared energy storage within multi-microgrids (MMGs). The model comprehensively considers system operation costs, energy storage allocation costs, and incorporates tiered carbon trading costs into the optimization objective, aiming to minimize overall system costs while maximizing carbon reduction benefits. Additionally, the Big-M method is employed to transform nonlinear constraints into a linear form, and case studies of specific MMGs systems are analyzed to validate the significant economic and environmental advantages of shared energy storage stations compared to individually allocated storage for each microgrid. Furthermore, the paper provides an in-depth discussion on the optimal configuration schemes, economic costs, and the impacts on carbon trading under varying scales of renewable energy integration. These findings offer theoretical insights for the optimal deployment of shared energy storage in the future development of power systems.

With the increasing penetration rate of new energy, the reliability and flexible operation ability of new power system are limited, and the application of demand response technology can effectively solve the contradiction of insufficient peak regulation of power grid. In order to solve the problem of how industrial flexible loads participate in system scheduling, this paper proposes an optimization method of industrial flexible coincidence scheduling based on probabilistic demand response potential assessment, and analyzes the typical industrial loads. Firstly, the basic idea of iterative adaptive clustering algorithm is introduced, and an improved fuzzy C-means clustering algorithm based on adaptive clustering center selection is proposed to generate typical daily load scenarios. Then, the load production characteristics of typical continuous and non-continuous production industries are studied, and their adjustable potential is analyzed. Secondly, the demand response potential evaluation method is established for two types of demand response types: peak cutting and valley filling. Finally, with the aim of minimizing the comprehensive operation cost of power system, the optimal scheduling model of industrial flexible load based on probabilistic demand response potential evaluation is established. Numerical results show that the proposed method is effective and reasonable.