Search Results (7,010)

Ensuring the safety and reliability of photovoltaic (PV) installations requires accurate electrical arc fault detection. This work presents a computational arc fault detection framework that combines fixed-length windowing, Fast Fourier Transform (FFT)-based features, and supervised machine learning classifiers. Data were acquired using an Arc Fault Circuit Interrupter (AFCI) test bench developed based on IEC 63027. Current and voltage signals were partitioned into 200-sample windows, DC-offset corrected, and Hann-windowed signals. Each window generated 204 statistical and spectral attributes used to train and evaluate Random Forest (RF), K-Nearest Neighbors (KNN), Multilayer Perceptron (MLP), and Convolutional Neural Network (CNN) models. Hyperparameters were tuned by grid search with TimeSeriesSplit cross-validation, comparing min–max normalization and Z–Score standardization. On a 15% hold-out test set, CNN with Z–Score achieved F1 = 0.9982 and recall = 0.9975, followed by MLP (F1 = 0.9957) and RF (F1 = 0.9821). Amortized per-window inference latencies were ≈ 0.0035 ms for RF, ≈ 0.0016 ms for MLP with Z–Score, and ≈ 0.032 ms for CNN. These classifier-stage timings indicate computational compatibility with edge-oriented implementation but do not constitute an end-to-end IEC 63027 AFCI compliance assessment. The framework targets integration into PV inverters at Mackenzie Presbyterian University’s solar plant. Full article

Deep networks dominate photovoltaic (PV) thermographic fault detection but need large annotated datasets and resist interpretation. We present a lightweight, interpretable infrared workflow combining support-vector-machine (SVM) module/background segmentation from four handcrafted features with an adaptive grid analysis labelling regions as nominal-intensity, high-intensity anomaly or low-intensity anomaly relative to a module-internal reference; the anomaly classes are inspection candidates, not confirmed faults. Evaluation used 21 close-range images of one 20 W module—recorded with the camera’s visible-light edge fusion active, so they are fused infrared/visible frames—and all 596 of a public five-sector UAV dataset. Segmentation against manual masks reached a mean intersection-over-union of 0.64; a feature ablation shows intensity statistics dominate, and an end-to-end Otsu pipeline gives almost the same high-intensity share (4.54% versus 4.50%): the SVM contributes reproducibility—removing the manual segmentation threshold, though not the empirical +48/−60 offsets—not accuracy. High-intensity regions concentrated in the module’s lower half, co-locating with a bus-bar defect known from hardware inspection—suggestive, not validated. The single-module, image-level close-range evaluation is optimistic, and the UAV shares, from a separately trained SVM, illustrate cross-domain application only. Segmentation runs at about 15 images per second on CPU. The method is a relative-intensity thermal screening workflow, not a validated defect-diagnosis or plant-health assessment method, and applies only where acquisition is controlled and the offsets are recalibrated for the target camera and palette. Full article

Accurate prediction of photovoltaic (PV) cell power from electroluminescence (EL) images is a key enabler for automated quality assessment and performance estimation in PV manufacturing and diagnostics. However, most existing image-based deep learning models are developed and evaluated for a single PV cell technology, limiting their ability to generalize across the wide variety of cell types used in practice. This work investigates the impact of PV cell technology on power prediction accuracy and proposes a novel Convolutional Hierarchical Mixture of Experts (CHME) architecture to overcome these generalization limitations. First, convolutional neural networks and feature-based machine learning models are evaluated on multiple PV cell technologies. While technology-specific convolutional models achieve low mean absolute errors (MAEs) of 0.02–0.04 when tested on the same cell type, their performance deteriorates substantially (MAEs of 0.08–0.23) when applied to different technologies. Feature-based models exhibit greater robustness across technologies but at the cost of lower prediction accuracy. To address these limitations, the proposed CHME model combines multiple pretrained technology-specific convolutional experts with a discriminator network that automatically identifies the PV cell technology and selects the most appropriate expert for power prediction. Experimental results demonstrate that CHME achieves the best overall performance, reducing the MAE to 0.0262 compared with 0.0298 for the best standalone convolutional model, while preserving adaptability to heterogeneous datasets. These results demonstrate that explicitly accounting for PV cell technology significantly improves image-based power prediction and that the proposed hierarchical mixture-of-experts framework provides an accurate, scalable, and easily retrainable solution for real-world PV diagnostic systems. Full article

Agriculture is increasingly challenged by water scarcity, climate change, and rising energy demands, requiring more efficient and sustainable irrigation solutions. Conventional irrigation systems often lack the capability to adapt their operation to changing soil conditions and renewable energy availability, resulting in inefficient water and energy use. This study proposes a digital twin-based framework for energy management and irrigation optimization in photovoltaic (PV)-powered smart agriculture systems using Internet of Things (IoT) soil monitoring. The proposed system integrates a physical irrigation infrastructure, an IoT monitoring network, a fuzzy logic control layer, and a digital twin environment that periodically synchronizes the virtual model with IoT measurements to support the system representation and decision-making. The digital twin models soil moisture, temperature, nutrient levels, PV energy generation, battery state of charge, and irrigation water consumption. The virtual representation was periodically aligned with the physical system using measurements transmitted through the long-range (LoRa)-based network. An energy-aware irrigation scheduling strategy was developed to optimize irrigation timing based on soil conditions, battery status, and solar energy availability. The framework was evaluated using field data collected in a real apple orchard through an ESP32-based IoT platform and a standalone PV-powered irrigation system; quantitative experimental validation was performed for the soil twin. The results demonstrate high soil twin synchronization accuracy, with an overall RMSE of 1.47 percentage points and R² of 0.981, based on experimental field measurements. The energy twin and irrigation twin were evaluated using experimentally acquired sensor data together with model-based performance assessment, demonstrating the potential of the proposed digital twin framework for integrated water–energy management in smart agriculture. Full article

Driven by low-carbon mandates, hybrid power vessels integrating diesel, battery, shore-power, and photovoltaic vessel (PV) systems are emerging as a key green shipping pathway. However, operation scheduling is essentially complex due to the integration between supply-side routing and load-side energy dispatch, which is compounded by multi-dimensional uncertainties in PV generation, port-grid loads, and feeder delays. To address this, we formulate a unified two-stage robust optimization model. The objective is to simultaneously minimize operating costs and enhance port-grid friendliness by coordinating on-board energy management and shore-power interactions. In detail, the first stage determines routing and sailing speeds, while the second stage allocates multi-source power under a worst-case budgeted polyhedral uncertainty set. A piecewise-linearization scheme handles the cubic speed–power relation, rendering a tractable mixed-integer linear programming problem. The problem is efficiently solved via a tailored Benders decomposition algorithm, utilizing a genetic-algorithm warm-start to substantially accelerate convergence. Validated on three real-world networks via 1000 Monte Carlo scenarios, the proposed model reduces mean operating costs by 20.2–23.5% and suppresses cost variance by over 60% compared to deterministic approaches. Full article

Driven by the rapid expansion of the new energy industry and the growing demand for photovoltaic (PV) power plant construction, optimizing site selection to ensure operational efficiency and stability has emerged as a critical imperative. Addressing the lack of precise zonation assessment systems for PV development in China’s arid and semi-arid regions, this study introduces a multi-dimensional, five-tier suitability evaluation framework. By integrating Boolean logic, the Bayesian Best–Worst Method (B-BWM), and Equal Interval classification, we developed a multi-resolution integrated assessment framework where spatial boundaries are mainly constrained by 30 m topographic and land-cover data. Results indicate that the candidate PV construction space (the upper two suitability classes, PDSI ≥ 3.40) spans 0.85 million km² concentrated in central-western Inner Mongolia and southeastern Xinjiang, and model validation achieves an Area Under the Curve (AUC) of 0.79. The annual technical potential reaches 19,069 TWh, equivalent to approximately 207% of China’s total electricity consumption in 2023. If fully developed, this potential offers a theoretical annual CO₂ emission reduction ranging from 12.69 to 21.48 billion tons under different conversion efficiency scenarios, with a baseline estimate of 14.65 billion tons. These multi-resolution integrated findings provide a useful spatial reference for preliminary site screening in arid and semi-arid regions, support China’s “Dual Carbon” goals, and offer a practical methodological approach for renewable energy planning on marginal lands. Full article

High photovoltaic (PV) penetration in distribution networks can cause significant voltage deviations, making effective voltage regulation a critical issue for distribution system operators. Step voltage regulators (SVRs) provide a cost-effective solution; however, their performance strongly depends on control parameter settings that should account for PV-induced voltage fluctuations. Because installing high-resolution sensors at all PV sites is impractical, this paper proposes a method for determining SVR control parameters using short-term PV output estimation. The proposed approach assumes that only large-capacity PV systems are equipped with high-resolution sensors and estimates the outputs of other PV systems using the inverse distance weighting (IDW) method. Based on historical power flow data and estimated PV outputs, the method determines 48 sets of SVR control parameters for the following day while capturing rapid voltage rises caused by PV fluctuations. Numerical simulations using a practical distribution network model developed from an actual Chubu Electric Power Grid Co., Inc. system demonstrate the effectiveness of the proposed method. The results show that the allowable PV hosting capacity increases by 25% compared with a method without short-term PV output estimation, while the increase in SVR tap operations remains within an acceptable range. Full article

The problem of load frequency control (LFC) becomes more severe with the extensive integration of photovoltaic (PV) generation owing to the intermittent nature of the source. In this study, a machine learning approach is developed to design the proportional–integral–derivative (PID) controller of a two-area PV–thermal LFC system, extending a prior proportional–integral (PI) benchmark to full PID action. A Random Forest model is trained to predict the relationship between the six PID gains and the closed-loop integral of time-multiplied absolute error (ITAE), yielding an accurate performance model (test R² = 0.933) that is subsequently searched by a metaheuristic optimizer to determine the controller gains; the resulting controller is termed ML-PID. The novelty of the approach lies in employing the learned model not as a controller or a physical-quantity predictor, as in existing ML-based LFC studies, but as a reusable performance model that maps the controller gains directly to the closed-loop index and guides the PID design. To isolate and quantify the contribution of the learned model, the same three optimizers, namely the Cheetah Optimizer (CO), the Grey Wolf Optimizer (GWO), and Particle Swarm Optimization (PSO), are also applied directly to the plant, yielding purely metaheuristic controllers (CO-PID, GWO-PID, and PSO-PID) that are compared against the machine learning-assisted designs under identical algorithms and computational budget, with CO selected on the basis of the Friedman and Wilcoxon tests. The proposed ML-PID-CO controller attains the minimum ITAE under a step-load disturbance, approximately 70% lower than that of the reference SCHO-PI controller and comparable to the directly optimized controllers, with reduced control effort. Under a simultaneous variation in the plant time constants, it is the most robust of all controllers, exhibiting the smallest Δf₁ undershoot and a performance that degrades about 4.2 times less than that of the reference. The results show that a learned performance model provides a good and reusable basis for PID design. It can be searched over repeatedly once built and reduces the per-design simulation burden relative to direct metaheuristic tuning, while the design is largely independent of the optimizer used. Full article

This study considers the multi-objective optimization of a standalone hybrid nanogrid system (HNGS) providing electricity power to a residential load in an off-grid area of Djelfa Province, Algeria. The focus of this study is to obtain the optimum design of a standalone hybrid nanogrid system consisting of photovoltaic (PV) panels, wind turbines (WTs), battery storage (BT), diesel generators (DGs), and power converters to satisfy the energy demand of residential consumers in Djelfa Province, Algeria. In this context, four multi-objective optimization algorithms (MOPs), NSGA-II, MOPSO, MOSSA, and MODE, are used to solve the optimal sizing problem of the proposed system. The formulated multi-objective optimization problem takes into account multiple performance criteria such as cost of energy (COE), loss of power supply probability (LPSP), renewable energy penetration, and diesel generator usage reduction, balancing economic, reliability, and sustainability aspects. The optimization process optimizes critical design parameters, including the size of the PV system, the number of wind turbines, and the size of the battery storage system, for a realistic operating scenario. The optimization algorithms are combined with an energy management strategy (EMS) that helps to coordinate the power flow distribution between various parts of the system to achieve optimum system performance. The effectiveness of each of the proposed approaches is analyzed based on the obtained results, where it was found that the MODE algorithm provides the best compromise solution, with a COE of 0.167 USD/kWh and an LPSP of 6.372%, and the lowest carbon dioxide emissions of 205.1 kg/year compared to MOPSO, NSGA-II, and MOSSA. Moreover, the results obtained from this process will provide a set of feasible design solutions, which will allow decision-makers to choose the most suitable design solution based on technical and economic specifications. Full article

With the advancement of China’s “dual carbon” goals and the green transformation of the railway sector, railway stations, as key energy-consuming nodes, require integrated energy systems that support low-carbon and renewable energy utilization. This review focuses on zero-carbon railway station integrated energy systems incorporating photovoltaic (PV) generation, energy storage, hydrogen systems, and charging facilities. Based on existing studies, the paper systematically reviews system configuration methods, operational strategies, and capacity optimization approaches. It first summarizes the roles of photovoltaic, energy storage, and hydrogen systems in railway station energy supply and outlines representative integration frameworks. It then compares standalone operation and coordinated multi-energy complementary operation, with particular attention to technical challenges in renewable energy accommodation, energy storage coordination, and hydrogen utilization. Mainstream capacity optimization approaches are further reviewed according to different energy configurations, including photovoltaic systems, energy storage systems (ESSs), hydrogen systems, and multi-energy complementary systems, with emphasis on optimization objectives, constraint formulations, and solution methodologies. The review shows that existing studies have gradually shifted from single-energy configurations toward coordinated multi-energy planning, but limitations remain in load forecasting accuracy, dynamic operational optimization, and large-scale engineering validation. Existing uncertainty management methods mainly include stochastic programming, robust optimization, chance-constrained optimization, and scenario-based approaches, which are used to address renewable energy fluctuations and load uncertainties. Future research should strengthen uncertainty modeling, real-time scheduling, and case study platforms considering diverse meteorological and load scenarios. This review provides a theoretical reference for planning and optimizing zero-carbon railway station integrated energy systems. Full article

Assessing the impacts of large-scale photovoltaic plants on soil properties is critical for sustainable land management in arid regions. This study examined a 50 MWp utility-scale PV facility in Siziwangqi, Inner Mongolia, northern China, a cold semi-arid desert steppe where renewable-energy development overlaps with fragile wind-eroded ecosystems. A spatially resolved sampling design contrasted soils at key micro-positions relative to the panels, namely Front, Under, and Behind, with adjacent natural Controls. By integrating laser diffraction analysis, multifractal modeling, and the EPIC erodibility equation, we evaluated soil particle-size probability redistribution and EPIC-estimated intrinsic erodibility. Compared with the silt-dominated surface soil of the natural steppe, soils within the photovoltaic plant exhibited fine-particle retention and lower information and correlation dimensions (D1 and D2), indicating stronger local clustering and a less even particle-size probability distribution. The EPIC-estimated erodibility factor decreased from 0.452 in the Control to approximately 0.361 in the PV micro-locations. These associations should be interpreted as texture- and SOC-based model estimates rather than direct measurements of wind erosion. Overall, multifractal parameters provide complementary proxy descriptors for detecting PV-induced particle sorting and potential changes in intrinsic soil erodibility, underscoring the need for field validation and adaptive management in dryland PV landscapes. These findings provide physical soil evidence for evaluating the environmental sustainability of dryland PV development and for supporting adaptive land management in solar farms. Full article

Photovoltaic (PV) power exhibits high variability and non-stationarity due to irradiance fluctuations, cloud shading, and seasonal changes, which complicate ultra-short-term multi-step forecasting. This study proposes a Solar-Historical Prior Network (SHPNet) for forecasting at 5 min resolution. SHPNet integrates a solar-geometry clear-sky prior-guided temporal convolutional network (SGCP-TCN), a similar historical window (SHW) branch, and horizon-wise adaptive fusion (HA). SGCP-TCN estimates clear-sky power potential from site coordinates and timestamps and reformulates direct power prediction as clear-sky power ratio forecasting. SHW retrieves training windows from the same intra-day time slot that exhibit similar power–irradiance evolution, thereby constructing a non-parametric historical prior, while HA determines horizon-specific fusion weights based on validation errors. Unlike purely data-driven predictors and conventional similar-day methods, SHPNet combines a physically interpretable power scale with input-window-level historical evolution patterns and adaptively balances the two priors across forecasting horizons. Across the two sites, SHPNet reduced the mean MAE and RMSE by 14.07% and 10.26%, respectively, compared with the original TCN, while increasing the mean R² from 0.8279 to 0.8613. Evaluations under different weather conditions and across seasons demonstrate consistent forecasting performance, while convergence analysis confirms stable training behavior. Full article

Rural houses in cold regions of China usually have high energy demands, particularly for space heating and domestic hot water. Passive design can reduce building energy demand, but additional renewable energy systems are still needed to improve on-site energy supply. This study evaluates the performance and limitations of photovoltaic (PV) and solar water heating (SWH) systems in a passive-designed rural house in Xi’an, China. Hourly simulations were conducted for PV-only and PV–battery configurations with different south-facing roof coverage ratios and battery capacities, together with an evacuated-tube SWH system. The results show that PV electricity supply was limited by the mismatch between household electricity demand and PV generation. Household demand mainly occurred in the morning and evening, whereas PV generation was concentrated around noon. The 13 m² PV case achieved approximately 11% electricity supply capacity with a utilization ratio of 62%, while increasing the PV area to 50 m² raised the supply capacity to only 15% and reduced the utilization ratio to 23%. With battery storage, the largest PV–battery configuration supplied 48% of annual household electricity demand, while the overall electricity utilization ratio was 73%, indicating a trade-off between household electricity self-supply and system utilization. The SWH system showed better applicability for domestic hot water supply, with an annual average hot water supply capacity of 60.2% and an average device efficiency of 43.5%, but its winter performance remained weak. These results indicate that PV and SWH are useful but insufficient solar energy strategies for passive-designed rural houses. PV is mainly constrained by daily time mismatch, while SWH is mainly constrained by seasonal climate variation. Full article

This study presents a gate-to-gate life cycle assessment (LCA) comparing the environmental impact of three torrefied municipal solid waste (MSW) energy sources (S1) solar photovoltaic (PV), (S2) grid electricity from India and (S3) biomass combustion. Experiments performed in a laboratory setting produced a yield of transitory MSW torrefaction of 28–32% at an input fuel energy of 2 kWh/kg at 200–300 °C for 30–60 min. The ReCiPe 2016 midpoints [Global Warming Potential (GWP); Human Toxicity Potential (HTP); Acidification Potential (AP); Particulate Matter Formation Potential (PMFP)] showed PV produced the least number of emissions (GWP = 0.0426 kg CO₂ equivalent; HTP = 0.00988 kg 1,4-DB equivalent), while grid power produced the greatest number of emissions (GWP = 2.2 kg CO₂ equivalent). Biomass produced intermediate results (GWP = 1.546 kg CO₂ equivalent). All sources had an average positive net energy ratio of approximately 2.78. The results indicate that integrating renewable energy sources significantly increases the environmental and social benefits of the torrefaction process. Additionally, the study provides a MS Excel-Based LCA framework to use when data are limited. Full article

The digitalization of photovoltaic (PV) systems can support sustainable energy deployment by improving operational efficiency, system visibility, and energy extraction. However, many existing Internet of Things (IoT)-enabled solutions address monitoring and maximum power point tracking (MPPT) separately or depend on proprietary platforms, remote cloud services, and relatively costly hardware, which may restrict their accessibility and replication in small-scale and resource-constrained applications. This study presents the implementation and laboratory-scale experimental evaluation of an edge-IoT architecture that integrates real-time PV monitoring, embedded adaptive MPPT control, local data management, and visualization using low-cost hardware and open-source software. The proposed architecture combines an ESP32 microcontroller with a Raspberry Pi (RPi) local server to enable environmental and electrical sensing, edge-based control, message queuing telemetry transport (MQTT) communication, local data storage, and interactive visualization through the open-source Node-RED, InfluxDB, and Grafana platforms. An adaptive perturb-and-observe (AP&O) algorithm is implemented on the ESP32 to dynamically adjust the duty cycle of a DC–DC boost converter in response to changing operating conditions. The system is experimentally evaluated using a PV test bench equipped with a custom boost converter and sensing modules measuring eleven electrical and environmental parameters. The architecture achieved an average communication latency of 193 ± 23 ms and an average MPPT efficiency of 97.3 ± 0.54%. It also provided a power gain of 0.7 ± 0.5% compared with the conventional fixed-step perturb-and-observe method. By combining local processing, open-source software, low-cost components, and integrated monitoring and control, the proposed system reduces dependence on external cloud infrastructure while supporting responsive and accessible PV energy management. These results demonstrate its potential as a replicable technological framework for improving the operational sustainability and digital management of small-scale PV installations. Full article