Open research questions in Solar Radiation and Photovoltaics

What the literature leaves open

• Inaccurate estimation of future PV generation can lead to biased revenue projections, misleading payback calculations, and elevated investment risk. The majority of existing studies evaluate forecasting performance primarily through statistical error metrics.

  COST ASSESSMENT FOR GRID-EXPORT PV SYSTEMS USING LONG-TERM FORECASTING: A PRE-INVESTMENT STUDY FOR BESS DEPLOYMENT · 2026 · DOI

• The non-linear nature of PV output. The high variability of meteorological conditions. The need for accurate and reliable short-term PV power forecasting.

  Enhancing Solar Power Forecasting Accuracy Using HMPCS and Machine Learning Techniques: An Applied Study · 2026 · DOI

• The random and intermittent properties of renewable sources impose a great challenge in the operation of the systems and energy management. The limited availability of data for training and testing deep learning models. The need for a more comprehensive system that incorporates multiple sources of energy.

  Deep learning-based time-series solar power prediction for automatic multisource energy generation systems · 2026 · DOI

• Existing purely data-driven deep learning algorithms lack physical interpretability and are prone to overfitting and prediction failure under non-stationary meteorological conditions. The large-scale integration of photovoltaic power poses a serious threat to the frequency stability and security of microgrids.

  Physics-Informed Temporal Convolutional Network for Ultra-Fast Short-Term PV Power Forecasting Mitigating Atmospheric Electromagnetic Extinction · 2026 · DOI

• The study uses a limited number of operational PV plants in Turkey. The framework may not be applicable to other regions with different meteorological conditions.

  A cluster-aware XAI-MCDM framework for province-level solar PV suitability screening in Turkey using real operational PV data · 2026 · DOI

• To apply the proposed framework to other regions with different meteorological conditions. To evaluate the performance of the framework using a larger dataset of operational PV plants. To adapt the framework for use in other renewable energy applications.

  A cluster-aware XAI-MCDM framework for province-level solar PV suitability screening in Turkey using real operational PV data · 2026 · DOI

• The intermittency of solar generation creates operational challenges such as grid congestion, curtailment, and unstable export profiles. There is a need for accurate short-term forecasting and optimization-driven scheduling to mitigate these issues.

  Forecast-Guided Economic Dispatch of BESS in Grid-Export-Only PV Systems · 2026 · DOI

• Spatial and seasonal variability in solar radiation. Limited data availability. Need for robust validation.

  EVALUATION DU POTENTIEL SOLAIRE DES REGIONS NATURELLES DE LA GUINEE · 2026 · DOI

• High accuracy, low computing time (1.2 s), scalable, real-time applicability, hybrid integration of advanced techniques Model complexity, data dependency, and requires fine-tuning for specific conditions KMKFC-GCA supports adaptive clustering under a wide variety of meteorological conditions, and as a result, the model generalised easily, and its results are understandable. Although all the methods that are incorporated in the hybrid framework, i.e., DOST, NARXNN, and KMKFC-GCA, are not new, the novelty of this framework lies in the optimization of these methods and their association. The special combination of DOST signal decomposition, NARXNN, temporal prediction, and KMKFC -GCA adaptive clustering is an efficient solution to the problem of predicting the solar radiation under different atmospheric conditions. Such practice not only enhances the accuracy of forecasting but also makes it more interpretable, which is a great breakthrough in comparison with single techniques and increases the predictability of solar energy in practical conditions. However, addressing a few limitations, such as optimizing the model for real-time applications to reduce computational complexity and improving robustness across varying weather conditions enhance the model’s applicability in diverse environments. The research organisation is as follows: Section 2 describes the methodology, a hybrid procedure for forecasting solar radiation; the proposed atmospheric clustering strategy is formulated in Sect. 3; Section 4 includes the experimental outcomes and assessments; and Sect. 5 summarises the findings and discusses future directions. 2 Methodology: Hybrid Solar Radiation Forecasting Framework This research presents a hybrid forecasting model that integrates advanced signal processing, deep learning, and clustering techniques to predict solar radiation under varying atmospheric conditions accurately. First, the DOST decomposes the solar radiation dataset into distinct frequency components that help to extract both the low-frequency and high-frequency patterns. Then, each decomposed sub-signal is forecast by the NARXNN, whose predictions are integrated through DOST principles into a single solar radiation forecast. In parallel, KMKFC-GCA classifies the data according to atmospheric transparency in the CI, enhancing the interpretability and performance in different weather conditions. This two-stage hybrid approach develops a robust solution with improved interpretability, which overcomes the challenges of non-linearity and forecast accuracy in solar radiation forecasting. Figure 1 illustrates the hybrid model of solar radiation prediction, which is a combination of several steps to determine the solar irradiance.

  Solar Radiation Forecasting Under Variable Atmospheric Conditions Using Signal Processing and Knowledge-Driven Clustering Approach · 2026 · DOI

• The study identifies a gap in current forecasting methodologies. The study highlights the need for innovative machine learning approaches for short-term PV power forecasting.

  Accurate Short-Term Photovoltaic (PV) Power Forecasting: Leveraging the Capabilities of AdaBoost and Decision Trees · 2026 · DOI

• Further studies can investigate the impact of air pollution parameters on solar radiation forecasting in other regions. The development of more advanced machine learning algorithms can improve forecasting accuracy.

  Air pollution and meteorological factors significantly enhance solar radiation forecasting accuracy in the western mediterranean region: a machine learning approach · 2026 · DOI

• The study identifies a gap in the existing literature on solar radiation forecasting, particularly in the context of Nigerian locations. The gap is addressed by developing a hybrid model that combines statistical and machine learning approaches.

  Day-Ahead Hourly Forecasting of Solar Radiation using a Physics-based Hybrid Machine Learning Model in Some Selected Locations in Nigeria · 2026 · DOI

• Competing land uses constrain evidence-based decision-making. Limited ground-based solar data is available. The deployment of photovoltaic infrastructure is fundamentally a spatial problem.

  Explainable geo-informatics for spatial solar suitability analysis in Gauteng Province, South Africa · 2026 · DOI

• Competing land uses and limited ground-based solar data constrain evidence-based decision-making in Gauteng Province, South Africa. The lack of spatially explicit, transparent, and sustainable approaches for identifying suitable renewable energy sites is a significant gap.

  Explainable geo-informatics for spatial solar suitability analysis in Gauteng Province, South Africa · 2026 · DOI

• Conventional mathematical models tend to fall short in estimating parameters accurately during abrupt variations in the atmospheric environment. Traditional analytical and numerical techniques often experience premature convergence and inaccurate results.

  Adaptive Hybrid Evolutionary Algorithms for High-Accuracy Solar Cell Modelling Under Dynamic Environmental Conditions · 2026 · DOI

• estimation, regarding and practical challenges. Future research should aim to address these issues by exploring broader applications and comparative analyses of various optimization strategies optimizing PV common share generalizability, to they specificity, implementation 3.Solar cell Modelling The Single Diode Model (SDM) is used to evaluate the performance of photovoltaic (PV) modules under various conditions, as manufacturer data alone is insufficient. This model includes five key parameters: photovoltaic current (Iph), reverse saturation current (I0), series resistance (Rs), shunt resistance (Rsh), and the ideality factor (n). Fig1. depicts the single diode solar cell model Fig1.Single diode solar cell model The PV cell is represented as an ideal solar cell with a current source in parallel to a diode. To model the PV cell, the first step is to determine the parameters based on specific temperature and irradiance conditions. The current-voltage (I-V) relationship can be expressed mathematically, and the junction thermal voltage (Vt) is calculated using known constants. The parameters are estimated by minimizing the difference between observed and calculated voltages in a PV string model. While the SDM is widely used, it has © 2026 The Author(s). Published by IJCOPE Journal. Website: https://ijcope.org/ 3 International Journal of Creative and Open Research in Engineering and Management ISSN: 3108-1754 (Online) Volume 02 Issue 05 May-2026 | Impact Factor: 3.5 limitations, particularly at low irradiance levels where recombination losses are not accounted. The I-V equation for double diode is as The first step of the modelling is to ascertain the parameters of the model equation for a specified temperature and irradiance value. Once these values are deduced, the solution of the equation can be found out using numerous numerical techniques.

  Adaptive Hybrid Evolutionary Algorithms for High-Accuracy Solar Cell Modelling Under Dynamic Environmental Conditions · 2026 · DOI

• The study identifies a gap in the existing literature in predicting PV output accurately. The study identifies a need for advanced predictive techniques to address the challenges in PV systems.

  Prediction of Photovoltaic Output using Single Candidate Optimizer – Artificial Neural Network · 2026 · DOI

• Additionally, further tuning or reconfiguration of the EP–ANN approach could be explored to enhance its generalization ability, though its current performance suggests it may be less suited for stable, cumulative- output prediction tasks compared to SCO–ANN.

  Prediction of Photovoltaic Output using Single Candidate Optimizer – Artificial Neural Network · 2026 · DOI

• The non-stationary nature of solar radiation. Variability in atmospheric conditions. The need for a model that can extrapolate to different climatic situations.

  Machine learning and deep learning based daily solar radiation forecasting for Fargo, North Dakota, USA · 2026 · DOI

• Traditional models often result in unreliable estimates due to the non-stationary nature of solar radiation. Hybrid systems, such as ANFIS, have been developed, but with limited performance.

  Machine learning and deep learning based daily solar radiation forecasting for Fargo, North Dakota, USA · 2026 · DOI

• The system relies on historical climatology rather than real-time meteorological inputs. The system does not integrate deep learning architectures. The system is limited to six Indian cities.

  S.U.R.A.J. (Solar Utility & Radiance Analytical Judgment): Localized ML-Based Forecasting for Diverse Indian Climates · 2026 · DOI

• Integration of real-time meteorological API feeds. Exploration of deep learning architectures such as LSTM and Temporal Fusion Transformer models.

  S.U.R.A.J. (Solar Utility & Radiance Analytical Judgment): Localized ML-Based Forecasting for Diverse Indian Climates · 2026 · DOI

• Further research can be conducted to improve the performance of the U-Net model. The application of deep learning models can be explored for other renewable energy sources.

  Evaluation of U-Net3+ and Attention U-Net for Solar Energy Estimation on Urban Rooftops using LiDAR DSMs · 2026 · DOI

• Traditional physical models are often computationally intensive and time-consuming. There is a need for a more efficient and accurate method for solar energy potential estimation.

  Evaluation of U-Net3+ and Attention U-Net for Solar Energy Estimation on Urban Rooftops using LiDAR DSMs · 2026 · DOI

• The selection of suitable sites for solar power plant installation is crucial for the overall success of clean energy projects. There is a need for a reliable and replicable model for sustainable solar energy planning and decision-making.

  An Integrated GIS–MCDM Framework Using BWM, SWARA, and MARCOS for Optimal Solar Power Plant Site Selection in Fars Province - Iran · 2026 · DOI

Most-cited papers in Solar Radiation and Photovoltaics

• The environmental factors affecting solar photovoltaic output · Renewable and Sustainable Energy Reviews · 2024 · 252 citations
• Improved multistep ahead photovoltaic power prediction model based on LSTM and self-attention with weather forecast data · Applied Energy · 2024 · 251 citations
• PV power forecasting based on data-driven models: a review · International Journal of Sustainable Engineering · 2021 · 147 citations
• Multi-step photovoltaic power forecasting using transformer and recurrent neural networks · Renewable and Sustainable Energy Reviews · 2024 · 124 citations
• An interpretable framework for modeling global solar radiation using tree-based ensemble machine learning and Shapley additive explanations methods · Applied Energy · 2024 · 107 citations
• Recent advances in intra-hour solar forecasting: A review of ground-based sky image methods · International Journal of Forecasting · 2022 · 101 citations
• A short-term forecasting method for photovoltaic power generation based on the TCN-ECANet-GRU hybrid model · Scientific Reports · 2024 · 90 citations
• Enhanced solar photovoltaic power prediction using diverse machine learning algorithms with hyperparameter optimization · Renewable and Sustainable Energy Reviews · 2024 · 90 citations
• Short-term photovoltaic power forecasting with feature extraction and attention mechanisms · Renewable Energy · 2024 · 86 citations
• A hybrid deep learning model with an optimal strategy based on improved VMD and transformer for short-term photovoltaic power forecasting · Energy · 2024 · 85 citations

Most recent work

• Accurate global horizontal irradiance estimation with hybrid convolutional neural network-informer-gated recurrent unit framework · Indian Journal of Physics · 2026
• AI for Solar PV Forecasting, MPPT, and Energy Management: A Comprehensive Review · International Journal of Computational and Electronic Aspects in Engineering · 2026
• AI-Based Automated Solar Farm Layout Optimization Using Satellite Imagery and Machine Learning · Zenodo (CERN European Organization for Nuclear Research) · 2026
• Physics-constrained multimodal vision transformer for ultra-short-term solar radiation forecasting error correction · Scientific Reports · 2026
• Comparative Analysis of Deep Learning Architectures for Short-Term Solar Power Forecasting · International Journal for Research in Applied Science and Engineering Technology · 2026
• Renewable energy management using explainable artificial intelligence · Frontiers in Energy Research · 2026
• Solar Radiation Forecasting Under Variable Atmospheric Conditions Using Signal Processing and Knowledge-Driven Clustering Approach · International Journal of Computational Intelligence Systems · 2026
• Accurate Short-Term Photovoltaic (PV) Power Forecasting: Leveraging the Capabilities of AdaBoost and Decision Trees · Journal of Electrical Engineering & Technology · 2026
• Air pollution and meteorological factors significantly enhance solar radiation forecasting accuracy in the western mediterranean region: a machine learning approach · Frontiers in Energy Research · 2026
• Day-Ahead Hourly Forecasting of Solar Radiation using a Physics-based Hybrid Machine Learning Model in Some Selected Locations in Nigeria · Nigerian Journal of Physics · 2026

Related topics in Computer Science