Search Results (201)

The exponential growth of machine learning (ML) applications in smart grid (SG) research over the past decade has generated a vast and fragmented body of literature that lacks systematic synthesis. This study addresses that gap by presenting a comprehensive bibliometric and science mapping analysis of the ML–smart grid (MLSG) research landscape to date, drawing on 4156 peer-reviewed publications indexed in the Elsevier Scopus database from 2009 to 2025. The principal contributions of this study are fourfold. First, it provides a rigorous quantitative mapping of MLSG publication growth from one document in 2009 to 1163 publications in 2025, representing a growth rate of 116,200%, thereby establishing a definitive baseline for tracking future scholarly development in the field. Second, it identifies the key actors driving MLSG research, including the most prolific authors (Nadeem Javaid, Alsabaan M.), leading institutions (King Saud University, Tennessee Technological University), and dominant nations (India, China, United States), which offers researchers and funding bodies actionable intelligence on collaboration opportunities and research leadership. Third, through keyword co-occurrence and cluster analysis, the study maps the three dominant thematic hotspots structuring current MLSG research—Smart Grid Security, Power Load Forecasting, and Advanced Energy Management—providing a structured intellectual framework that can guide future research prioritization. Fourth, the study delivers a critical thematic literature review of these three hotspots, synthesizing the most impactful ML methodologies and applications reported across 4156 publications, including deep learning-based intrusion detection, ensemble forecasting models, and reinforcement learning-driven energy management. Collectively, these contributions offer a robust evidence base for researchers, policymakers, and industry practitioners seeking to navigate, benchmark, and advance the field of ML-enabled smart grid systems. Full article

This study addresses the multi-objective optimization of characterizing a flat glass processing plant. To assess the operational conditions required for a flat glass processing small and medium-sized enterprise (SME) to become a prosumer compatible with renewable energy community (REC) participation. This work is motivated by the presence of more than 300 SMEs in Italy, like this, where RECs represent one of the few viable strategies for achieving the European Union’s 2050 decarbonization targets. The research is carried out in two scenarios; Scenario-I includes Stage-i and Stage-ii with the mutual goal of forecasting and optimizing. Forecasting is used in Stage-i to optimize the factory load, and in Stage-ii to shift and curtail energy loads based on the forecast, considering the Italian national energy price and the regional price bands (“fasce orarie”) F1, F2, and F3. Forecasting and the indicators of environmental and social performance are the means to ensure the best energy utilization and management, as they prove that the reduction in CO₂ emissions and benefits on the community level can be both obtainable. Subsequently, the techno-economic analysis and evaluation of prosumer-readiness conditions are carried out through the optimization of industrial energy demand: three optimization objectives are assessed in this study (i) energy cost, (ii) carbon emission, and (iii) load curtailment. Four algorithms are put into effect to solve the tri-objective optimization: multi-objective particle swarm optimization (MOPSO), multi-objective ant nesting algorithm (MOANA), non-dominated sorting genetic algorithm (NSGA-II), and multi-objective grey wolf optimization (MOGWO). The algorithms are validated in Stage-ii to find the desired optimum in the cost of energy, reduce peak formation, and carbon emissions. To achieve this goal, a stochastic approach based on Monte Carlo simulations and VIKOR is used to optimally select the results. The findings show that the NSGA-II, MOPSO, and MOANA are more effective in solving the problem, while the MOGWO algorithm more quickly finds the optimal solution. Based on the defined objectives, a new configuration for the energy community is introduced, together with a community well-being index and an evaluation of the resulting benefits for the factory. In Scenario-II, the PV plants’ installation on the factory is sized, and the excess energy shared with the grid is evaluated. The Scenario-II results show that 497.184 MWh (33.9%) of energy is shared with the grid. Both results suggest how optimized industrial demand profiles improve SME participation in future RECs. Full article

Air pollution negatively impacts human health and environmental sustainability, particularly in areas with high industrial activity. This study comparatively evaluated deep learning-based models for estimating PM₁₀ and SO₂ pollutants in Dilovası and Ereğli (Turkey), industrial areas with high pollutant loads. The study utilized Recurrent Neural Networks (RNNs), Long Short-Term Memory (LSTM), Gated Recurrent Units (GRUs), an RNN–GRU stacked hybrid model, an attention-based hybrid model, and the proposed GRU–LSTM dual-path hybrid model. The proposed method consists of four main stages: data conversion into a time-series format, data preprocessing and feature generation, model architecture development, and model training and performance evaluation. The dataset consisted of 365 daily PM₁₀ and SO₂ observations obtained from the Air Monitoring Center for the Dilovası and Ereğli monitoring stations. Model performance was evaluated using the coefficient of determination (R²), training time, root mean squared error (RMSE), mean squared error (MSE), and mean absolute error (MAE) metrics. The findings showed that the hybrid models provided higher accuracy compared to the single-track models. Specifically, the proposed GRU–LSTM dual-path hybrid model produced the highest R² and lowest error values for both pollutant parameters in both the Dilovası and Ereğli regions. In Dilovası, this model achieved R² = 0.97 for SO₂ and R² = 0.96 for PM₁₀; in Ereğli, it reached R² = 0.92 for SO₂ and R² = 0.98 for PM₁₀. Thus, it has been shown that the GRU–LSTM dual-path hybrid model, which models short-term and long-term temporal dependencies in parallel, is an effective and reliable method for air pollutant forecasting in industrial areas. These findings demonstrate the potential of the proposed model to support air quality monitoring, early warning systems, and environmental decision-making in industrial regions. Full article

Under the national context of energy transition and energy conservation, accurate prediction of thermal load in oil and gas gathering and transportation stations is crucial for ensuring operational safety and reducing energy consumption. To address the limitations of traditional forecasting methods in handling the nonlinear, non-stationary, and long-term temporal dependencies of thermal load data, this paper proposes a hybrid deep learning model that integrates convolutional neural networks (CNNs), long short-term memory (LSTM) networks, and an attention mechanism, namely the CNN–LSTM–Attention model. First, key influencing factors such as ambient temperature, return water temperature, and the previous hour’s thermal load were selected as model inputs through correlation analysis. Subsequently, a CNN was employed to extract spatial features from multi-source data, LSTM to capture temporal dependencies, and an attention mechanism to dynamically focus on critical operational nodes, thereby enhancing the model’s ability to perceive important features. The experimental results show that the proposed model performs excellently in heat load prediction, achieving a root mean square error of 5.98, a mean absolute error of 4.66, and a mean absolute percentage error of 9.66%, with an R-squared (R²) value of 0.9568. Its prediction accuracy and stability are significantly superior to those of the standalone CNN and standalone LSTM models. This study provides an effective algorithmic solution for precise thermal load forecasting in oil and gas gathering and transportation stations and offers insights for optimizing the applicability of deep learning models in industrial scenarios. Full article

Accurate building performance forecasting is critical for the design and renovation of energy-saving structures, but existing methods face four key challenges: heterogeneous data fusion (sensor streams, design parameters, and environmental sequences), non-stationary physical time series, model interpretability, and sample efficiency (e.g., limited commissioning data). To address these challenges, this paper proposes Fusion-Grounded Forecasting (FGF), which is a framework integrating a gated adaptive fusion layer, deterministic trend-season decomposition, an additive predictor with component decomposition, and Bayesian regularization. This framework is designed for next-hour forecasting broadcast to hourly resolution using hourly sensor data and monthly design parameters. The dataset covers 36 months (approximately 25,920 h). In addition to the combination of existing modules, the novelty lies in the integrated architecture, in which interpretable constraints can adjust the fusion layer in both directions, with decomposition prediction alignment supporting component attributes. The framework is verified on a proprietary 36-month dataset from institutional buildings using standard prediction metrics (MAE, RMSE, MAPE, and directional accuracy) and ablation studies for comparison against 10 baselines: SARIMAX, GPR, LSTM, XGBoost, N-HiTS, Informer, Autoformer, NAM, a physics-informed hybrid, and TFT. FGF achieves a 3.1% MAPE and 92.5% directional accuracy in hourly cooling load forecasting. Ablation confirmed the contribution of each module: removing gated fusion increased the MAPE to 6.8%. Compared with manual feature engineering, the speed of the framework is increased by 1680 times, and the cost is reduced by 99.6%. The explanatory index (counterfactual reliability: 0.95; Stability of functional importance: 0.11) is in compliance with audit requirements. These results indicate that FGF connects descriptive physics with quantitative prediction. However, this study is limited to a single institutional building; transferability to residential, commercial, or industrial buildings requires further verification. While waiting for this verification, FGF has demonstrated its potential as a transparent and efficient tool to build performance models. Full article

The review presents a comprehensive review of the influence of power quality indicators and operating conditions at industrial enterprises on the technical condition and reliability of electrical equipment. Harmonic distortion, voltage fluctuations and sags, load surges, overvoltages, and voltage unbalance are considered factors that increase thermal, electrical, and mechanical stresses in transformers, induction motors, cable lines, and overhead power lines. It is shown that these disturbances can increase RMS currents, additional losses, hot-spot temperature, vibration, and insulation aging rate, reducing equipment service life and increasing failure probability. The review links power quality disturbances with thermal aging models, remaining useful life assessment, and probabilistic reliability models, including the Weibull distribution. It is established that a correct remaining service life assessment requires considering not only individual disturbances but also the combined influence of voltage and current quality, load conditions, ambient temperature, and humidity. Particular attention is paid to modern monitoring and forecasting technologies, including IoT systems, multi-agent models, machine learning, and predictive diagnostics. These technologies enable the transition from scheduled maintenance to continuous multiparameter monitoring. A structure for quantitative risk assessment and practical recommendations for predictive maintenance of industrial electrical equipment are proposed. Full article

Against the backdrop of the global advancement of carbon neutrality goals and the energy transition in the building sector, zero-carbon buildings have emerged as pivotal enablers for achieving carbon neutrality in the construction industry. The rule-based scheduling of energy storage systems (ESS) is critical to enhancing energy efficiency and economic performance of buildings. This study takes the Jinan Zero-Carbon Operation Center Project in Shandong Province as the research object, developing a comprehensive technical framework covering the entire process from design to operation, and investigates the rule-based design and ESS scheduling strategies in response to Shandong’s newly implemented seasonal time-of-use (TOU) electricity pricing policy. First, core performance indicators are defined in accordance with national evaluation standards for zero-carbon buildings. Hourly building energy loads and photovoltaic (PV) generation profiles are simulated over a full year, which serves as the basis for determining the optimal PV installed capacity and ESS sizing. Second, an ESS scheduling strategy integrating PV generation forecasting and the seasonal TOU electricity price structure is formulated, with clear charging and discharging logic defined. Finally, the operational and economic performance of different scheduling modes are evaluated and compared through case studies. The results show that the annual PV generation ratio reaches 101.38%, with a self-consumption rate of 73% and a self-sufficiency rate of 72%, all meeting the core requirements for zero-carbon buildings. Compared with the conventional real-time scheduling mode (Mode 1), the proposed optimized mode (Mode 2) that incorporates TOU pricing and PV forecasting achieves an annual operational cost saving of 367,349 CNY, corresponding to a reduction of 47.02%. Distinct seasonal variations in core indicators are also observed: the PV generation ratio is lower in summer and winter but the self-consumption rate is higher, with the opposite trend in spring and autumn. The proposed technical framework and scheduling strategy provide practical guidance for the design and operational optimization of zero-carbon buildings and offer decision-making support for ESS operation under TOU electricity pricing policies. Full article

Accurate prediction of superheated steam temperature (SST) is critical for the safe and efficient operation of large-scale thermal power units, particularly under large load variations and high thermal inertia. This study proposes an iTransformer-based SST prediction framework (iTransformer-SST) to address limitations of conventional proportional–integral–derivative (PID) control and existing data-driven models in capturing multivariable coupling, time-delay effects, and physical consistency. Using the A-side subsystem of a 1000 MW thermal power unit, 19-dimensional process data were collected continuously over two months with a sampling interval of 2.4 s. After data preprocessing, time-lagged cross-correlation (TLCC) analysis combined with expert knowledge was employed for feature screening, resulting in ten highly relevant input variables. To enhance predictive robustness, the baseline iTransformer was augmented with a Local Temporal Convolution (LTC) module for local disturbance modeling and a physics-guided regularization term to enforce delayed monotonicity and smoothness constraints. In 240 min rolling forecasts of the final-stage superheater outlet temperature, the proposed model achieved a mean squared error (MSE) of 0.0887, a mean absolute error (MAE) of 0.2312, and a coefficient of determination ($R^2$) of 0.9650, significantly outperforming long short-term memory (LSTM), Informer, and the baseline iTransformer. The combined LTC and physics-guided design reduced MSE by 13.5%, demonstrating strong potential for feedforward-assisted SST control in industrial thermal power applications. Full article

Coal-fired power units remain integral to electricity supply in many regions while facing increasingly stringent environmental expectations. Bridging reliable generation with sustainability requires more than end-of-pipe controls; it demands continuous intelligence embedded in plant operations. This study introduces an industry-oriented monitoring framework that transforms historical operational records into actionable foresight, enabling on-the-fly orchestration of combustion conditions to anticipate sulfur oxide (SOx) concentrations. Leveraging 919 empirical data points collected in 2019 from Unit 8 of the Taichung Thermal Power Plant, the framework integrates robust data governance, targeted feature curation, and a neural network-based analytics core. Eight process variables—sulfur content, coal feed rate, fixed carbon, grinding rate, calorific value, excess air, air flow, and boiler efficiency—emerge as the most influential drivers through systematic selection and feature importance attribution. The resulting forecasting module exhibits near-perfect alignment with observed emissions (R² = 0.99), enabling near-real-time guidance for setpoint adjustments and facilitating compliance strategies under varying load and fuel-quality conditions. Beyond accuracy, the system is architected for scalability and portability, aligning with Industry 4.0 paradigms by coupling continuous sensing, data-driven decision support, and stakeholder transparency. By reframing emission oversight as a proactive, intelligent service rather than a static reporting function, the proposed approach advances operational resilience, regulatory compliance, and community trust, with direct implications for resource efficiency and circular economy initiatives across heavy industry. The framework reduces potential SOx emissions and improves energy utilization efficiency under varying operational conditions. This approach contributes to environmental sustainability by enabling proactive emission reduction and cleaner production practices. It supports regulatory compliance and aligns with global sustainability goals, including SDG 7 and SDG 13. Full article

Accurate multi-step building load forecasting is critical for optimizing energy management in smart grids and reducing operational costs. However, existing forecasting methods struggle with complex temporal dependencies, seasonal variations, and robust performance under noisy conditions. This paper proposes R-SATNet (Robust Self-Attention Transformer Network), a novel deep learning architecture that integrates multi-head self-attention mechanisms with robust optimization techniques for enhanced building load prediction. The proposed framework incorporates temporal feature extraction modules, adaptive noise suppression layers, and multi-scale attention blocks to capture both short-term fluctuations and long-term seasonal patterns. Extensive experiments on real-world building load datasets demonstrate that R-SATNet achieves superior forecasting accuracy with 15.7% lower RMSE and 12.3% improved MAPE compared to state-of-the-art methods. The model maintains robust performance under various noise conditions and provides reliable multi-step predictions up to 24 h ahead, making it highly suitable for practical smart energy system deployments. The proposed framework is validated across six diverse building datasets spanning commercial, residential, industrial, campus, mixed-use, and healthcare facilities, confirming its generalizability and practical applicability in heterogeneous smart energy environments. Full article

Under the dual pressures of global carbon emission reduction and production cost control, energy-intensive industrial enterprises are in urgent need of a balanced low-carbon operation strategy that reconciles economic benefits, environmental performance and production continuity. To address the limitations of existing methods in multi-dimensional objective balancing, this paper proposes a dynamic multi-objective optimization framework for industrial electricity consumption, integrating high-precision load forecasting and optimal scheduling. For load forecasting, an improved dual-gate optimization temporal attention long short-term memory (DGO-TA-LSTM) model is developed, which is modeled based on the one-year hourly electricity operation data (8760 samples) of a high-energy industrial enterprise in southern China, and its performance is verified via three standard metrics—the mean absolute error (MAE), root mean square error (RMSE) and mean absolute percentage error (MAPE)—compared with five mainstream baseline models. On this basis, when taking time-varying electricity-carbon factors and time-of-use electricity prices as dual guiding signals, a three-objective optimization model minimizing electricity cost, carbon emissions and load deviation is constructed, which is solved by the Non-Dominated Sorting Genetic Algorithm II (NSGA-II), with the Improved Gray Target Decision-Making (IGTD) method introduced to select the optimal compromise solution. Case study results show that the proposed scheme achieved a 1.9% reduction in electricity cost and a 30% reduction in carbon emissions compared with the unoptimized strategy, providing a feasible and scalable low-carbon operation path for industrial enterprises. Full article

Long-term time series forecasting (LTSF) is critical for modern power systems, energy management, and grid planning. Yet virtually all existing forecasting models employ stationary activation functions that apply identical nonlinear mappings regardless of temporal context—a fundamental mismatch with real-world load data, which exhibits strongly regime-dependent dynamics such as summer demand peaks, winter heating patterns, and overnight low-load periods. We address this gap by proposing TC-KAN (Time-Conditioned Kolmogorov–Arnold Network), the first forecasting architecture to augment KAN activation functions with position-aware coefficient parameterisation. The core innovation replaces the static polynomial coefficients in standard KAN activations with position-conditioned coefficients produced by a lightweight positional-embedding MLP, providing additional learnable capacity beyond standard KAN while adding negligible parameter overhead. TC-KAN further integrates a dual-pathway processing block—combining depthwise convolution for local temporal pattern extraction with the time-conditioned KAN layer for enhanced nonlinear transformation—within a channel-independent framework with Reversible Instance Normalisation. Experiments were conducted on four standard ETT benchmark datasets and the high-dimensional Weather dataset. TC-KAN achieves superior or competitive accuracy in most configurations while requiring merely 51K parameters—approximately 40% of DLinear and ∼100× fewer than iTransformer. On ETTh2, TC-KAN reduces the mean squared error by up to 61.4% over DLinear, and matches the current state-of-the-art iTransformer on ETTm2 at a fraction of the computational cost. This extreme parameter reduction circumvents the steep memory bottlenecks endemic to massive Transformer models, positioning TC-KAN as a highly practical architecture tailored precisely for resource-constrained edge deployments—such as on-device load forecasting inside smart grid sensors and industrial IoT controllers. Full article

Against the backdrop of the global energy transition toward deep electrification, the natural gas industry faces challenges, including increased load forecasting uncertainty and frequent extreme weather impacts. To enhance natural gas load forecasting accuracy and support system resilience planning, this study constructs a forecasting model based on quadratic decomposition and hybrid deep learning, incorporating an electricity substitution coefficient to characterize the coupling substitution effect between electricity and natural gas. Under the basic scenario, the VMD-WPD-TCN-BiGRU model is proposed. It employs variational mode decomposition and wavelet packet denoising for secondary signal denoising, combined with a time-series convolutional network and bidirectional gated recurrent unit to extract temporal features. Experiments demonstrate that, compared to mainstream methods such as CNN, BiLSTM, SVM, and XGBoost, this model achieves statistically significant reductions in MSE (11.11–96.21%), MAE (0.89–76.50%), and MAPE (4.10–67.94%), significantly improving forecasting accuracy. In the deep electrification scenario, the introduction of the electricity substitution coefficient further optimizes peak load forecasting for system pressure days under extreme low temperatures, elevating the overall R² to 0.9905 in the deep electrification scenario. Research indicates that the proposed model not only effectively improves the accuracy of short-term natural gas load forecasting but also provides quantitative support for enterprises to plan peak-shaving facilities, optimize pipeline networks, and respond to extreme weather emergencies in data silo environments. This contributes to strengthening the adaptability and long-term resilience of natural gas systems during the energy transition, thereby supporting the sustainable development of energy infrastructure. Full article

Energy management systems operating under dynamic electricity pricing require fast and cost-optimal control strategies for flexible loads. Mixed-integer linear programming (MILP) can compute theoretically optimal control trajectories but is computationally expensive and typically relies on accurate load forecasts, limiting its practical real-time applicability. This paper proposes a supervised imitation learning (IL) framework that learns optimal setpoint trajectories for a conventional proportional (P) controller directly from electricity price signals and temporal features, thereby eliminating the need for explicit load forecasting. The learned model predicts setpoint trajectories in an open-loop manner, while a lower-level P controller ensures stable closed-loop operation within a two-stage control architecture. The approach is validated in an industrial case study involving load shifting of a refrigeration system under dynamic electricity pricing and benchmarked against MILP optimization, reinforcement learning (RL), heuristic strategies, and various machine learning models. The MILP solution achieves a cost reduction of 21.07% and represents a theoretical upper bound under perfect information. The proposed Transformer model closely approximates this optimum, achieving 19.33% cost reduction while enabling real-time inference. Overall, the results demonstrate that the proposed supervised IL approach can achieve near-optimal control performance with substantially reduced computational effort for real-time energy management applications. Full article

To address the operational challenges of industrial microgrids under different weather conditions, this study proposes an optimal dispatch strategy for energy storage systems under representative weather scenarios. Photovoltaic (PV) power generation is first forecast using a Light Gradient Boosting Machine (LightGBM) model, while the load input is prepared based on recent historical demand patterns, and the forecasting performance is evaluated under representative sunny and cloudy scenarios. A mathematical microgrid model incorporating PV generation, battery energy storage, load demand, and grid interaction is then established, in which the total operating cost is minimized subject to time-of-use electricity pricing, battery degradation, and state-of-charge (SOC) constraints. Based on this formulation, an optimization-based day-ahead scheduling strategy is implemented. Under the selected representative sunny and cloudy conditions, the proposed method reduced the daily operating cost by 19.93% and 4.44%, respectively. Over seven representative days, the average cost reduction rate reached 12.54%, thereby confirming its economic effectiveness under representative weather scenarios. Full article