Search Results (964)

Floods stand as one of the most harmful natural disasters, which have become more dangerous because of climate change effects on urban structures and agricultural fields. This research presents a comprehensive flood mapping approach that combines multi-sensor satellite data with a machine learning method to evaluate the July 2021 flood in the Netherlands. The research developed 25 different feature scenarios through the combination of Sentinel-1, Landsat-8, and Radarsat-2 imagery data by using backscattering coefficients together with optical Normalized Difference Water Index (NDWI) and Hue, Saturation, and Value (HSV) images and Synthetic Aperture Radar (SAR)-derived Grey Level Co-occurrence Matrix (GLCM) texture features. The Random Forest (RF) classifier was optimized before its application based on two different flood-prone regions, which included Zutphen’s urban area and Heijen’s agricultural land. Results demonstrated that the multi-sensor fusion scenarios (S18, S20, and S25) achieved the highest classification performance, with overall accuracy reaching 96.4% (Kappa = 0.906–0.949) in Zutphen and 87.5% (Kappa = 0.754–0.833) in Heijen. For the flood class F1 scores of all scenarios, they varied from 0.742 to 0.969 in Zutphen and from 0.626 to 0.969 in Heijen. Eventually, the addition of SAR texture metrics enhanced flood boundary identification throughout both urban and agricultural settings. Radarsat-2 provided limited benefits to the overall results, since Sentinel-1 and Landsat-8 data proved more effective despite being freely available. This study demonstrates that using SAR and optical features together with texture information creates a powerful and expandable flood mapping system, and RF classification performs well in diverse landscape settings. Full article

Traditional unmanned aerial vehicle (UAV) detection and tracking methods have long faced the twin challenges of high cost and poor efficiency. In real-world battlefield environments with complex backgrounds, occlusions, and varying speeds, existing techniques struggle to track small UAVs accurately and stably. To tackle these issues, this paper presents an enhanced YOLOv8N-Drone-based algorithm for improved target tracking of small UAVs. Firstly, a novel module named C2f-DSFEM (Depthwise-Separable and Sobel Feature Enhancement Module) is designed, integrating Sobel convolution with depthwise separable convolution across layers. Edge detail extraction and multi-scale feature representation are synchronized through a bidirectional feature enhancement mechanism, and the discriminability of target features in complex backgrounds is thus significantly enhanced. For the feature confusion problem, the improved lightweight Context Anchored Attention (CAA) mechanism is integrated into the Neck network, which effectively improves the system’s adaptability to complex scenes. By employing a position-aware weight allocation strategy, this approach enables adaptive suppression of background interference and precise focus on the target region, thereby improving localization accuracy. At the level of loss function optimization, the traditional classification loss is replaced by the focal loss (Focal Loss). This mechanism effectively suppresses the contribution of easy-to-classify samples through a dynamic weight adjustment strategy, while significantly increasing the priority of difficult samples in the training process. The class imbalance that exists between the positive and negative samples is then significantly mitigated. Experimental results show the enhanced YOLOv8 boosts mean average precision (Map@0.5) by 12.3%, hitting 99.2%. In terms of tracking performance, the proposed YOLOv8 N-Drone algorithm achieves a 19.2% improvement in Multiple Object Tracking Accuracy (MOTA) under complex multi-scenario conditions. Additionally, the IDF1 score increases by 6.8%, and the number of ID switches is reduced by 85.2%, indicating significant improvements in both accuracy and stability of UAV tracking. Compared to other mainstream algorithms, the proposed improved method demonstrates significant advantages in tracking performance, offering a more effective and reliable solution for small-target tracking tasks in UAV applications. Full article

Privacy-sensitive electricity scene classification requires robust models under data localization constraints, making federated learning (FL) a suitable framework. Existing FL frameworks face two critical challenges in multi-label electricity scene classification: (1) Label correlations and their strengths significantly impact classification performance. (2) Electricity scene data and labels show distributional inconsistencies across regions. However, current FL frameworks lack explicit modeling of label correlation strengths, and locally trained regional models naturally capture these differences, leading to regional differences in their model parameters. In this scenario, the server’s standard single-stage aggregation often over-averages the global model’s parameters, reducing its discriminative ability. To address these issues, we propose FMMAN, a federated multi-stage attention neural network for multi-label electricity scene classification. The main contributions of this FMMAN lie in label correlation learning and the stepwise model aggregation. It splits the client–server interaction into multiple stages: (1) Clients train models locally to encode features and label correlation strengths after receiving the server’s initial model. (2) The server clusters these locally trained models into K groups to ensure that models within a group have more consistent parameters and generates K prototype models via intra-group aggregation to reduce over-averaging. The K models are then distributed back to the clients. (3) Clients refine their models using the K prototypes with contrastive group-specific consistency regularization to further mitigate over-averaging, and sends the refined model back to the server. (4) Finally, the server aggregates the models into a global model. Experiments on multi-label benchmarks verify that FMMAN outperforms baseline methods. Full article

The mining industry (MI) in mineral-rich regions is pivotal for economic growth but is challenged by significant pollution and emissions. This study examines Guangxi, a representative region in China, in light of the country’s “Dual Carbon” goals. We quantified carbon emissions from the MI from 2005 to 2021, employing the generalized Divisia index method (GDIM) to analyze the factors driving these emissions. Additionally, a system dynamics (SD) model was developed, integrating economic, demographic, energy, environmental, and policy variables to assess decarbonization strategies and the potential for carbon decoupling. The key findings include the following: (1) Carbon accounting analysis reveals a rising emission trend in Guangxi’s MI, predominantly driven by electricity consumption, with the non-ferrous metal mining sector contributing the largest share of total emissions. (2) The primary drivers of carbon emissions were identified as economic scale, population intensity, and energy intensity, with periodic fluctuations in sector-specific drivers necessitating coordinated policy adjustments. (3) Scenario analysis showed that the Emission Reduction Scenario (ERS) is the only approach that achieves a carbon peak before 2030, indicating that it is the most effective decarbonization pathway. (4) Between 2022 and 2035, carbon decoupling from total output value is projected to improve under both the Energy-Saving Scenario (ESS) and ERS, achieving strong decoupling, while the resource extraction shows limited decoupling effects often displaying an expansionary connection. This study aims to enhance the understanding and promote the advancement of green and low-carbon development within the MI in mineral-rich regions. Full article

Bare land exposed by glacier retreat provides new opportunities for ecosystem development. Investigating primary vegetation succession in deglaciated regions can provide significant insights for ecological restoration, particularly for future climate change scenarios. Nonetheless, research on this topic in the Qinghai–Tibet Plateau has been exceedingly limited. This study aimed to investigate vegetation succession in the deglaciated area of the Zepu glacier during the Little Ice Age in southeastern Tibet. Quadrat surveys were performed on arboreal communities, and trends in vegetation change were assessed utilizing multi-year (1986–2024) remote sensing data. The findings indicate that vegetation succession in the Zepu glacier deglaciated area typically adheres to a sequence of bare land–shrub–tree, divided into four stages: (1) shrub (species include Larix griffithii Mast., Hippophae rhamnoides subsp. yunnanensis Rousi, Betula utilis D. Don, and Populus pseudoglauca C. Wang & P. Y. Fu); (2) broadleaf forest primarily dominated by Hippophae rhamnoides subsp. yunnanensis Rousi; (3) mixed coniferous–broadleaf forest with Hippophae rhamnoides subsp. yunnanensis Rousi and Populus pseudoglauca C. Wang & P. Y. Fu as the dominant species; and (4) mixed coniferous–broadleaf forest dominated by Picea likiangensis (Franch.) E. Pritz. Soil depth and NDVI both increase with succession. Species diversity is significantly higher in the third stage compared to other successional stages. In addition, soil moisture content is significantly greater in the broadleaf-dominated communities than in the conifer-dominated communities. An analysis of NDVI from 1986 to 2024 reveals an overall positive trend in vegetation recovery in the area, with 93% of the area showing significant vegetation increase. Temperature is the primary controlling factor for this recovery, showing a positive correlation with vegetation cover. The results indicate that Key ecological indicators—including species composition, diversity, NDVI, soil depth, and soil moisture content—exhibit stage-specific patterns, reflecting distinct phases of primary succession. These findings enhance our comprehension of vegetation succession in deglaciated areas and their influencing factors in deglaciated areas, providing theoretical support for vegetation restoration in climate change. Full article

This article presents a comprehensive assessment of mudflow risk in the Talgar River basin through the application of Multi-Criteria Decision Analysis (MCDA) methods and numerical modeling using the Rapid Mass Movement Simulation (RAMMS) environment. The first part of the study involves a spatial assessment of mudflow hazard and susceptibility using GIS technologies and MCDA. The key condition for evaluating mudflow hazard is the identification of factors influencing the formation of mudflows. The susceptibility assessment was based on viewing the area as an object of spatial and functional analysis, enabling determination of its susceptibility to mudflow impacts across geomorphological zones: initiation, transformation, and accumulation. Relevant criteria were selected for analysis, each assigned weights based on expert judgment and the Analytic Hierarchy Process (AHP). The results include maps of potential mudflow hazard and susceptibility, showing areas of hazard occurrence and risk impact zones within the Talgar River basin. According to the mudflow hazard map, more than 50% of the basin area is classified as having a moderate hazard level, while 28.4% is subject to high hazard, and only 1.8% falls under the very high hazard category. The remaining areas are categorized as very low (4.1%) and low (14.7%) hazard zones. In terms of susceptibility to mudflows, 40.1% of the territory is exposed to a high level of susceptibility, 35.6% to a moderate level, and 5.5% to a very high level. The remaining areas are classified as very low (1.8%) and low (15.6%) susceptibility zones. The predictive performance was evaluated through Receiver Operating Characteristic (ROC) curves, and the Area Under the Curve (AUC) value of the mudflow hazard assessment is 0.86, which indicates good adaptability and relatively high accuracy, while the AUC value for assessing the susceptibility of the territory is 0.71, which means that the accuracy of assessing the susceptibility of territories to mudflows is within the acceptable level of model accuracy. To refine the spatial risk assessment, mudflow modeling was conducted under three scenarios of glacial-moraine lake outburst using the RAMMS model. For each scenario, key flow parameters—height and velocity—were identified, forming the basis for classification of zones by impact intensity. The integration of MCDA and RAMMS results produced a final mudflow risk map reflecting both the likelihood of occurrence and the extent of potential damage. The presented approach demonstrates the effectiveness of combining GIS analysis, MCDA, and physically-based modeling for comprehensive natural hazard assessment and can be applied to other mountainous regions with high mudflow activity. Full article

To address the significant degradation of image visibility and contrast in turbid media, this paper proposes an enhanced image dehazing algorithm. Unlike traditional polarimetric dehazing methods that exclusively attribute polarization information to airlight, our approach integrates object radiance polarization and airlight polarization for haze removal. First, sky regions are localized through multi-scale fusion of polarization and intensity segmentation maps. Second, region-specific transmittance estimation is performed by differentiating haze-occluded regions from haze-free regions. Finally, target radiance is solved using boundary constraints derived from non-haze regions. Compared with other dehazing algorithms, the method proposed in this paper demonstrates greater adaptability across diverse scenarios. It achieves higher-quality restoration of targets with results that more closely resemble natural appearances, avoiding noticeable distortion. Not only does it deliver excellent dehazing performance for land fog scenes, but it also effectively handles maritime fog environments. Full article

Land use/land cover (LULC) change constitutes a critical driver influencing regional carbon cycling processes. Optimizing LULC structures represents a significant pathway toward the realization of carbon neutrality. This study takes Liaoning Province as a case area to analyze LULC changes from 2000 to 2020 and to assess their impacts on land use carbon emissions (LUCE) and ecosystem carbon storage (ECS). To accelerate the achievement of carbon neutrality, four development scenarios are established: natural development (ND), low-carbon emission (LCE), high-carbon storage (HCS), and carbon neutrality (CN). For each scenario, corresponding optimization objectives and constraint conditions are defined, and a multi-objective LULC optimization coupling model is formulated to optimize both the quantity structure and spatial pattern of LULC. On this basis, the model quantifies ECS and LUCE under the four scenarios and evaluates the economic value of each scenario and its contribution to the carbon neutrality target. Results indicate the following: (1) From 2000 to 2020, the extensive expansion of construction land resulted in a reduction in ECS by 12.72 × 10⁶ t and an increase in LUCE by 150.44 × 10⁶ t; (2) Compared to the ND scenario, the LCE scenario exhibited the most significant performance in controlling carbon emissions, while the HCS scenario achieved the highest increase in carbon sequestration. The CN scenario showed significant advantages in reducing LUCE, enhancing ECS, and promoting economic growth, achieving a reduction of 0.18 × 10⁶ t in LUCE, an increase of 118.84 × 10⁶ t in ECS, and an economic value gain of 3386.21 × 10⁶ yuan. This study optimizes the LULC structure from the perspective of balancing economic development, LUCE reduction, and ECS enhancement. It addresses the inherent conflict between regional economic growth and ecological conservation, providing scientific evidence and policy insights for promoting LULC optimization and advancing carbon neutrality in similar regions. Full article

Refining the land use structure can boost land utilization efficiency and curtail regional carbon emissions. Nevertheless, prior research has predominantly concentrated on static linear planning analysis. It has failed to account for how future dynamic alterations in driving factors (such as GDP and population) affect simulation outcomes and how the land use spatial configuration impacts the attainment of the carbon-neutrality goal. In this research, 1 km spatial resolution LULC products were employed to meticulously simulate multiple land use scenarios across China at the national level from 2030 to 2060. This was performed by taking into account the dynamic changes in driving factors. Subsequently, an analysis was carried out on the low-carbon land use spatial structure required to reach the carbon-neutrality target. The findings are as follows: (1) When employing the PLUS (Patch—based Land Use Simulation) model to conduct simulations of various land use scenarios in China by taking into account the dynamic alterations in driving factors, a high degree of precision was attained across diverse scenarios. The sustainable development scenario demonstrated the best performance, with kappa, OA, and FoM values of 0.9101, 93.15%, and 0.3895, respectively. This implies that the simulation approach based on dynamic factors is highly suitable for national-scale applications. (2) The simulation accuracy of the PLUS and GeoSOS-FLUS (Systems for Geographical Modeling and Optimization, Simulation of Future Land Utilization) models was validated for six scenarios by extrapolating the trends of influencing factors. Moreover, a set of scenarios was added to each model as a control group without extrapolation. The present research demonstrated that projecting the trends of factors having an impact notably improved the simulation precision of both the PLUS and GeoSOS-FLUS models. When contrasted with the GeoSOS-FLUS model, the PLUS model attained superior simulation accuracy across all six scenarios. The highest precision indicators were observed in the sustainable development scenario, with kappa, OA, and FoM values reaching 0.9101, 93.15%, and 0.3895, respectively. The precise simulation method of the PLUS model, which considers the dynamic changes in influencing factors, is highly applicable at the national scale. (3) Under the sustainable development scenario, it is anticipated that China’s land use carbon emissions will reach their peak in 2030 and achieve the carbon-neutrality target by 2060. Net carbon emissions are expected to decline by 14.36% compared to the 2020 levels. From the perspective of dynamic changes in influencing factors, the PLUS model was used to accurately simulate China’s future land use. Based on these simulations, multi-scenario predictions of future carbon emissions were made, and the results uncover the spatiotemporal evolution characteristics of China’s carbon emissions. This study aims to offer a solid scientific basis for policy-making related to China’s low-carbon economy and high-quality development. It also intends to present Chinese solutions and key paths for achieving carbon peak and carbon neutrality. Full article

This study investigates changes in Near-Surface Air Temperature (NSAT) over the South African region using weather station data, reanalysis products, and Coupled Model Intercomparison Project Phase 6 (CMIP6) model outputs. It is shown that, based on ERA5 reanalysis, the average NSAT increase in the region (45–10° S, 0–50° E) for the period 1940–2023 was 0.11 ± 0.04 °C. Weak multi-decadal changes in NSAT were observed from 1940 to the mid-1970s, followed by a rapid warming trend starting in the mid-1970s. Weather station data generally confirm these results, although they exhibit considerable inter-station variability. An ensemble of 33 CMIP6 models also reproduces these multi-decadal NSAT change characteristics. Specifically, the average model-simulated NSAT values for the region increased by 0.63 ± 0.12 °C between the periods 1940–1969 and 1994–2023. Based on the results of the comparison between weather station observations, reanalysis, and models, we utilize projections of NSAT changes from the analyzed ensemble of 33 CMIP6 models until the end of the 21st century under various Shared Socioeconomic Pathway (SSP) scenarios. These projections indicate that the average NSAT of the South African region will increase between 1994–2023 and 2070–2099 by 0.92 ± 0.36 °C under the SSP1-2.6 scenario, by 1.73 ± 0.44 °C under SSP2-4.5, by 2.52 ± 0.50 °C under SSP3-7.0, and by 3.17 ± 0.68 °C under SSP5-8.5. Between 1994–2023 and 2025–2054, the increase in average NSAT for the studied region, considering inter-model spread, will be 0.49–1.15 °C, depending on the SSP scenario. Furthermore, climate warming in South Africa, both in the next 30 years and by the end of the 21st century, is projected to occur according to all 33 CMIP6 models under all considered SSP scenarios. The main spatial feature of this warming is a more significant increase in NSAT over the landmass of the studied region compared to its surrounding waters, due to the stabilizing role of the ocean. Full article

Corydalis yanhusuo W. T. Wang, commonly known as Yanhusuo, is an important and rare medicinal plant resource in China. Its habitat integrity is facing severe challenges due to climate change and human activities. Establishing an integrative quality zoning system for this species is of significant practical importance for resource conservation and adaptive management. This study integrates multiple data sources, including 121 valid distribution points, 37 environmental factors, future climate scenarios (SSP126 and SSP585 pathways for the 2050s and 2090s), and measured content of tetrahydropalmatine (THP) from 22 sampling sites. A predictive framework for habitat suitability and spatial distribution of effective components was constructed using a multi-model coupling approach (MaxEnt, ArcGIS spatial analysis, and co-kriging method). The results indicate that the MaxEnt model exhibits high prediction accuracy (AUC > 0.9), with the dominant environmental factors being the precipitation of the wettest quarter (404.8~654.5 mm) and the annual average temperature (11.8~17.4 °C). Under current climatic conditions, areas of high suitability are concentrated in parts of Central and Eastern China, including the Sichuan Basin, the middle–lower Yangtze plains, and coastal areas of Shandong and Liaoning. In future climate scenarios, the center of suitable areas is predicted to shift northwestward. The content of THP is significantly correlated with the mean diurnal temperature range, temperature seasonality, and the mean temperature of the wettest quarter (p < 0.01). A comprehensive assessment identifies the Yangtze River Delta region, Central China, and parts of the Loess Plateau as the optimal integrative quality zones. This research provides a scientific basis and decision-making support for the sustainable utilization of C. yanhusuo and other rare medicinal plants in China. Full article

As China’s first systematic assessment of high-risk Amaranthaceae invaders, this study addresses a critical knowledge gap identified in the National Invasive Species Inventory, in which four invasive Amaranthaceae species (Dysphania ambrosioides, Celosia argentea, Amaranthus palmeri, and Amaranthus spinosus) are prioritized due to CNY 2.6 billion annual ecosystem damages in China. By coupling multi-species comparative analysis with a parameter-optimized Maximum Entropy (MaxEnt) model integrating climate, soil, and topographical variables in China under Shared Socioeconomic Pathways (SSP) 126/245/585 scenarios, we reveal divergent expansion mechanisms (e.g., 247 km faster northward shift in A. palmeri than D. ambrosioides) that redefine invasion corridors in the North China Plain. Under current conditions, the suitable habitats of these species span from 92° E to 129° E and 18° N to 49° N, with high-risk zones concentrated in central and southern China, including the Yunnan–Guizhou–Sichuan region and the North China Plain. Temperature variables (Bio: Bioclimatic Variables; Bio6, Bio11) were the primary contributors based on permutation importance (e.g., Bio11 explained 56.4% for C. argentea), while altitude (e.g., 27.3% for A. palmeri) and UV-B (e.g., 16.2% for A. palmeri) exerted lower influence. Model validation confirmed high accuracy (mean area under the curve (AUC) > 0.86 and true skill statistic (TSS) > 0.6). By the 2090s, all species showed net habitat expansion overall, although D. ambrosioides exhibited net total contractions during mid-century under the SSP126/245 scenarios, C. argentea experienced reduced total suitability during the 2050s–2070s despite high-suitability growth, and A. palmeri and A. spinosus expanded significantly in both total and highly suitable habitat. All species shifted their distribution centroids northward, aligning with warming trends. Overall, these findings highlight the critical role of temperature in driving range dynamics and underscore the need for latitude-specific monitoring strategies to mitigate invasion risks, providing a scientific basis for adaptive management under global climate change. Full article

There is an obvious correlation between the photovoltaic (PV) output of different physical levels; that is, the overall power change trend of large-scale regional (high-level) stations can provide a reference for the prediction of the output of sub-regional (low-level) stations. The current PV prediction methods have not deeply explored the multi-level PV power generation elements and have not considered the correlation between different levels, resulting in the inability to obtain potential information on PV power generation. Moreover, traditional probabilistic prediction models lack adaptability, which can lead to a decrease in prediction performance under different PV prediction scenarios. Therefore, a probabilistic prediction method for short-term PV power based on multi-level adaptive dynamic integration is proposed in this paper. Firstly, an analysis is conducted on the multi-level PV power stations together with the influence of the trend of high-level PV power generation on the forecast of low-level power generation. Then, the PV data are decomposed into multiple layers using the complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) and analyzed by combining fuzzy entropy (FE) and mutual information (MI). After that, a new multi-level model prediction method, namely, the improved dual dynamic adaptive stacked generalization (I-Stacking) ensemble learning model, is proposed to construct short-term PV power generation prediction models. Finally, an improved dynamic adaptive kernel density estimation (KDE) method for prediction errors is proposed, which optimizes the performance of the prediction intervals (PIs) through variable bandwidth. Through comparative experiments and analysis using traditional methods, the effectiveness of the proposed method is verified. Full article

This study proposes an integrated, machine learning-based multi-objective optimization framework to evaluate and optimize the utilization of steel slag in road base layers, simultaneously addressing economic costs and environmental impacts. A comprehensive dataset of 482 scenarios was engineered based on literature-informed parameters, encompassing transport distance, processing energy intensity, initial moisture content, gradation adjustments, and regional electricity emission factors. Four advanced tree-based ensemble regression algorithms—Random Forest Regressor (RFR), Extremely Randomized Trees (ERTs), Gradient Boosted Regressor (GBR), and Extreme Gradient Boosting Regressor (XGBR)—were rigorously evaluated. Among these, GBR demonstrated superior predictive performance (R² > 0.95, RMSE < 7.5), effectively capturing complex nonlinear interactions inherent in slag processing and logistics operations. Feature importance analysis via SHapley Additive exPlanations (SHAP) provided interpretative insights, highlighting transport distance and energy intensity as dominant factors affecting unit cost, while moisture content and grid emission factor predominantly influenced CO₂ emissions. Subsequently, the Gradient Boosted Regressor model was integrated into a Non-Dominated Sorting Genetic Algorithm II (NSGA-II) framework to explore optimal trade-offs between cost and emissions. The resulting Pareto front revealed a diverse solution space, with significant nonlinear trade-offs between economic efficiency and environmental performance, clearly identifying strategic inflection points. To facilitate actionable decision-making, the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) method was applied, identifying an optimal balanced solution characterized by a transport distance of 47 km, energy intensity of 1.21 kWh/ton, moisture content of 6.2%, moderate gradation adjustment, and a grid CO₂ factor of 0.47 kg CO₂/kWh. This scenario offered a substantial reduction (45%) in CO₂ emissions relative to cost-minimized solutions, with a moderate increase (33%) in total cost, presenting a realistic and balanced pathway for sustainable infrastructure practices. Overall, this study introduces a robust, scalable, and interpretable optimization framework, providing valuable methodological advancements for sustainable decision making in infrastructure planning and circular economy initiatives. Full article

As a key enabling technology for energy transition, the smart grid is propelling the global power system to evolve toward greater efficiency, reliability, and sustainability. Based on the three-dimensional analysis framework of “technology–policy–application”, this study systematically sorts out the technical architecture, regional development mode, and typical application scenarios of the smart grid, revealing the multi-dimensional challenges that it faces. By using the methods of literature review, cross-national case comparison, and technology–policy collaborative analysis, the differentiated paths of China, the United States, and Europe in the development of smart grids are compared, aiming to promote the integration and development of smart grid technologies. From a technical perspective, this paper proposes a collaborative framework comprising the perception layer, network layer, and decision-making layer. Additionally, it analyzes the integration pathways of critical technologies, including sensors, communication protocols, and artificial intelligence. At the policy level, by comparing the differentiated characteristics in policy orientation and market mechanisms among China, the United States, and Europe, the complementarity between government-led and market-driven approaches is pointed out. At the application level, this study validates the practical value of smart grids in optimizing energy management, enhancing power supply reliability, and promoting renewable energy consumption through case analyses in urban smart energy systems, rural electrification, and industrial sectors. Further research indicates that insufficient technical standardization, data security risks, and the lack of policy coordination are the core bottlenecks restricting the large-scale development of smart grids. This paper proposes that a new type of intelligent and resilient power system needs to be constructed through technological innovation, policy coordination, and international cooperation, providing theoretical references and practical paths for energy transition. Full article