Hour-Ahead Photovoltaic Power Prediction Combining BiLSTM and Bayesian Optimization Algorithm, with Bootstrap Resampling for Interval Predictions.

Photovoltaic (PV) power prediction plays a critical role amid the accelerating adoption of renewable energy sources. This paper introduces a bidirectional long short-term memory (BiLSTM) deep learning (DL) model designed for forecasting photovoltaic power one hour ahead. The dataset under examination originates from a small PV installation located at the Polytechnic School of the University of Alcala. To improve the quality of historical data and optimize model performance, a robust data preprocessing algorithm is implemented. The BiLSTM model is synergistically combined with a Bayesian optimization algorithm (BOA) to fine-tune its primary hyperparameters, thereby enhancing its predictive efficacy. The performance of the proposed model is evaluated across diverse meteorological and seasonal conditions. In deterministic forecasting, the findings indicate its superiority over alternative models employed in this research domain, specifically a multilayer perceptron (MLP) neural network model and a random forest (RF) ensemble model. Compared with the MLP and RF reference models, the proposed model achieves reductions in the normalized mean absolute error (nMAE) of 75.03% and 77.01%, respectively, demonstrating its effectiveness in this type of prediction. Moreover, interval prediction utilizing the bootstrap resampling method is conducted, with the acquired prediction intervals carefully adjusted to meet the desired confidence levels, thereby enhancing the robustness and flexibility of the predictions.

A dataset for photovoltaic energy potential of the Yaqui Valley, Sonora, México.

The adequate records of climatological variables, especially temperature and radiation, constitute the basis to be able to determine the photovoltaic and agrivoltaic potential of a region, for this purpose, the data collected must be extremely precise, so it is required that they be focused, current and reliable, to have an adequate estimation. This paper presents the dataset used to estimate the photovoltaic potential of the Yaqui Valley, Sonora, México, for agrivoltaic systems, with the objective of determining the photovoltaic energy generation capacity. Specific records of temperature, radiation and humidity variables obtained from 21 meteorological stations distributed in the Yaqui Valley are used to determine the photovoltaic potential in relation to planting surfaces and commercially available agrivoltaic technologies. To do this, the data was filtered, grouped, and normalized according to the ranges of the variables required for the analysis, this data comprise the last three years.

Solar photovoltaic panel production in Mexico: A novel machine learning approach.

This study examines the potential for widespread solar photovoltaic panel production in Mexico and emphasizes the country's unique qualities that position it as a strong manufacturing candidate in this field. An advanced model based on artificial neural networks has been developed to predict solar photovoltaic panel plant metrics. This model integrates a state-of-the-art non-linear programming framework using Pyomo as well as an innovative optimization and machine learning toolkit library. This approach creates surrogate models for individual photovoltaic plants including production timelines. While this research, conducted through extensive simulations and meticulous computations, unveiled that Latin America has been significantly underrepresented in the production of silicon, wafers, cells, and modules within the global market; it also demonstrates the substantial potential of scaling up photovoltaic panel production in Mexico, leading to significant economic, social, and environmental benefits. By hyperparameter optimization, an outstanding and competitive artificial neural network model has been developed with a coefficient of determination values above 0.99 for all output variables. It has been found that water and energy consumption during PV panel production is remarkable. However, water consumption (33.16 × 10-4 m3/kWh) and the emissions generated (1.12 × 10-6 TonCO2/kWh) during energy production are significantly lower than those of conventional power plants. Notably, the results highlight a positive economic trend, with module production plants generating the highest profits (35.7%) among all production stages, while polycrystalline silicon production plants yield comparatively lower earnings (13.0%). Furthermore, this study underscores a critical factor in the photovoltaic panel production process which is that cell production plants contribute the most to energy consumption (39.7%) due to their intricate multi-stage processes. The blending of Machine Learning and optimization models heralds a new era in resource allocation for a more sustainable renewable energy sector, offering a brighter, greener future.

IoT System Based on Artificial Intelligence for Hot Spot Detection in Photovoltaic Modules for a Wide Range of Irradiances.

Infrared thermography (IRT) is a technique used to diagnose Photovoltaic (PV) installations to detect sub-optimal conditions. The increase of PV installations in smart cities has generated the search for technology that improves the use of IRT, which requires irradiance conditions to be greater than 700 W/m2, making it impossible to use at times when irradiance goes under that value. This project presents an IoT platform working on artificial intelligence (AI) which automatically detects hot spots in PV modules by analyzing the temperature differentials between modules exposed to irradiances greater than 300 W/m2. For this purpose, two AI (Deep learning and machine learning) were trained and tested in a real PV installation where hot spots were induced. The system was able to detect hot spots with a sensitivity of 0.995 and an accuracy of 0.923 under dirty, short-circuited, and partially shaded conditions. This project differs from others because it proposes an alternative to facilitate the implementation of diagnostics with IRT and evaluates the real temperatures of PV modules, which represents a potential economic saving for PV installation managers and inspectors.

A comprehensive analysis of charge transfer effects on donor-pyrene (bridge)-acceptor systems using different substituents.

The alternant polycyclic aromatic hydrocarbon pyrene has photophysical properties that can be tuned with different donor and acceptor substituents. Recently, a D (donor)-Pyrene (bridge)-A (acceptor) system, DPA, with the electron donor N,N-dimethylaniline (DMA), and the electron acceptor trifluoromethylphenyl (TFM), was investigated by means of time-resolved spectroscopic measurements (J. Phys. Chem. Lett. 2021, 12, 2226-2231). DPA shows great promise for potential applications in organic electronic devices. In this work, we used the ab initio second-order algebraic diagrammatic construction method ADC(2) to investigate the excited-state properties of a series of analogous DPA systems, including the originally synthesized DPAs. The additionally investigated substituents were amino, fluorine, and methoxy as donors and nitrile and nitro groups as acceptors. The focus of this work was on characterizing the lowest excited singlet states regarding charge transfer (CT) and local excitation (LE) characters. For the DMA-pyrene-TFM system, the ADC(2) calculations show two initial electronic states relevant for interpreting the photodynamics. The bright S1 state is locally excited within the pyrene moiety, and an S2 state is localized ~0.5 eV above S1 and characterized as a donor to pyrene CT state. HOMO and LUMO energies were employed to assess the efficiency of the DPA compounds for organic photovoltaics (OPVs). HOMO-LUMO and optical gaps were used to estimate power conversion and light-harvesting efficiencies for practical applications in organic solar cells. Considering the systems using smaller D/A substituents, compounds with the strong acceptor NO2 substituent group show enhanced CT and promising properties for use in OPVs. Some of the other compounds with small substituents are also found to be competitive in this regard.

c-Si PV module recycling: Analysis of the use of a mechanical pre-treatment to reduce the environmental impact of thermal treatment and enhance materials recovery.

The current increase in the use of photovoltaic (PV) energy demands the search for solutions to recycle end-of-life modules. This study evaluated the use of a mechanical pre-treatment in the thermal recycling of c-Si crystalline PV modules, which were submitted to recycling routes to separate and concentrate the materials of interest. The first route was constituted by only thermal treatment, and the second route was constituted by a mechanical pre-treatment to remove the polymers from the backsheet, and subsequent thermal treatment. The exclusively thermal route was performed at 500°C, varying dwell times between 30 and 120 minutes in the furnace. In this route, the best results were obtained in 90 minutes, with a maximum degradation of 68% of the polymeric mass. In route 2, a micro-grinder rotary tool was used to remove the polymers from the backsheet and, subsequently, thermal treatment performed at 500°C, with dwell times in the furnace ranging between 5 and 30 minutes. The mechanical pre-treatment removed about 10.32 ± 0.92% of the mass of the laminate PV module. By this route, only 20 minutes of thermal treatment were needed for the total decomposition of the polymers, that is, a reduction of 78% in the oven time. With route 2, it was possible to obtain a concentrate with 30 times more silver than the PV laminate and 40 times more than a high-concentration ore. Furthermore, with route 2 it was possible to reduce the environmental impact of heat treatment and energy consumption.

Photovoltaic Glass Waste Recycling in the Development of Glass Substrates for Photovoltaic Applications.

Because of the increasing demand for photovoltaic energy and the generation of end-of-life photovoltaic waste forecast, the feasibility to produce glass substrates for photovoltaic application by recycling photovoltaic glass waste (PVWG) material was analyzed. PVWG was recovered from photovoltaic house roof panels for developing windows glass substrates; PVWG was used as the main material mixed with other industrial waste materials (wSG). The glass was casted by air quenching, annealed, and polished to obtain transparent substrates samples. Fluorine-doped tin oxide (FTO) was deposited as back contact on the glass substrates by spray pyrolysis. The chemical composition of the glass materials was evaluated by X-ray fluorescence (XRF), the thermal stability was measured by differential thermal analysis (DTA) and the transmittance was determined by UV-VIS spectroscopy. The surface of the glass substrates and the deposited FTO were observed by scanning electron microscopy (SEM), the amorphous or crystalline state of the specimens were determined by X-ray diffraction (XRD) and the sheet resistance was evaluated by the four-point probe method. The sheet resistance of the deposited FTO on the wSG substrate was 7.84 ± 3.11 Ω/□, lower than that deposited on commercial soda-lime glass (8.48 ± 3.67 Ω/□), meaning that this material could present improved conduction of the produced electrons by the photovoltaic effect. This process may represent an alternative to produce glass substrates from waste materials that could be destined for photovoltaic applications, especially the production of ecological photovoltaic windows.

Power Disturbance Monitoring through Techniques for Novelty Detection on Wind Power and Photovoltaic Generation.

Novelty detection is a statistical method that verifies new or unknown data, determines whether these data are inliers (within the norm) or outliers (outside the norm), and can be used, for example, in developing classification strategies in machine learning systems for industrial applications. To this end, two types of energy that have evolved over time are solar photovoltaic and wind power generation. Some organizations around the world have developed energy quality standards to avoid known electric disturbances; however, their detection is still a challenge. In this work, several techniques for novelty detection are implemented to detect different electric anomalies (disturbances), which are k-nearest neighbors, Gaussian mixture models, one-class support vector machines, self-organizing maps, stacked autoencoders, and isolation forests. These techniques are applied to signals from real power quality environments of renewable energy systems such as solar photovoltaic and wind power generation. The power disturbances that will be analyzed are considered in the standard IEEE-1159, such as sag, oscillatory transient, flicker, and a condition outside the standard attributed to meteorological conditions. The contribution of the work consists of the development of a methodology based on six techniques for novelty detection of power disturbances, under known and unknown conditions, over real signals in the power quality assessment. The merit of the methodology is a set of techniques that allow to obtain the best performance of each one under different conditions, which constitutes an important contribution to the renewable energy systems.

Outdoor and synthetic performance data for PV devices concerning the weather conditions and capacitor values of I-V tracer.

This article presents the outdoor and synthetic performance data concerning the main electrical parameters estimated from the I-V curve for three photovoltaic technologies (HIT, m-Si and CIGS) and the weather conditions (irradiance, ambient and panel temperature). Synthetic data were generated by simulating in OpenModelica software the impact of weather conditions on device performance, considering an irradiance range between 50 and 1300 W/m2. The outdoor data corresponds to the performance of the evaluated PV modules in outdoor tests in Medellin-Colombia for ten months using capacitive I-V tracers. In both cases, different capacitor values were considered to evaluate the effect on the I-V curve behavior of devices.

How Does Neural Network Model Capacity Affect Photovoltaic Power Prediction? A Study Case.

The use of models capable of forecasting the production of photovoltaic (PV) energy is essential to guarantee the best possible integration of this energy source into traditional distribution grids. Long Short-Term Memory networks (LSTMs) are commonly used for this purpose, but their use may not be the better option due to their great computational complexity and slower inference and training time. Thus, in this work, we seek to evaluate the use of neural networks MLPs (Multilayer Perceptron), Recurrent Neural Networks (RNNs), and LSTMs, for the forecast of 5 min of photovoltaic energy production. Each iteration of the predictions uses the last 120 min of data collected from the PV system (power, irradiation, and PV cell temperature), measured from 2019 to mid-2022 in Maceió (Brazil). In addition, Bayesian hyperparameters optimization was used to obtain the best of each model and compare them on an equal footing. Results showed that the MLP performs satisfactorily, requiring much less time to train and forecast, indicating that they can be a better option when dealing with a very short-term forecast in specific contexts, for example, in systems with little computational resources.

Construction and characterization of organic photovoltaic cells sensitized by Chrysanthemum based natural dye.

In this work, natural dyes from three different species of the same flower family (Chrysanthemum), which containing anthocyanin were extracted and properly prepared to be used as photosensitizers in DSSCs construction. The cells were fabricated with titanium dioxide nanoparticles (TiO2) for the photoanodes, whereas platinum electrodes were used for the photocathodes. To understand the behavior of light absorption in addition to the coloring components present in the dyes and the molecular functional groups present in the samples, the UV-Vis absorption spectroscopy and FTIR spectroscopy were used respectively. The performance and efficiency of solar cells were evaluated to establish the photovoltaic criteria for each DSSC built. Through electrochemical characterizations, it was possible to notice that the highest photovoltaic conversion efficiency was obtained with the Chrysanthemum Violet (CV) cell, with efficiency (η) of 1.348%, compared to 1.229% and 0.485% for the Chrysanthemum Green (CG) and Chrysanthemum Blue (CB) cells, respectively. The CV cell also has the highest open circuit voltage (VOC) at 0.58 V. The results corroborate to present the organic solar cells as a viable option for the electric energy generation.

Application of AI for Short-Term PV Generation Forecast.

The efficient use of the photovoltaic power requires a good estimation of the PV generation. That is why the use of good techniques for forecast is necessary. In this research paper, Long Short-Term Memory, Bidirectional Long Short-Term Memory and the Temporal convolutional network are studied in depth to forecast the photovoltaic power, voltage and efficiency of a 1320 Wp amorphous plant installed in the Technology Support Centre in the University Rey Juan Carlos, Madrid (Spain). The accuracy of these techniques are compared using experimental data along one year, applying 1 timestep or 15 min and 96 step times or 24 h, showing that TCN exhibits outstanding performance, compared with the two other techniques. For instance, it presents better results in all forecast variables and both forecast horizons, achieving an overall Mean Squared Error (MSE) of 0.0024 for 15 min forecasts and 0.0058 for 24 h forecasts. In addition, the sensitivity analyses for the TCN technique is performed and shows that the accuracy is reduced as the forecast horizon increases and that the 6 months of dataset is sufficient to obtain an adequate result with an MSE value of 0.0080 and a coefficient of determination of 0.90 in the worst scenarios (24 h of forecast).

Photovoltaic Electrochemically Driven Degradation of Calcon Dye with Simultaneous Green Hydrogen Production.

In this study, for the first time, the production of green hydrogen gas (H2) in the cathodic compartment, in concomitance with the electrochemical oxidation (EO) of an aqueous solution containing Calcon dye at the anodic compartment, was studied in a PEM-type electrochemical cell driven by a photovoltaic (PV) energy source. EO of Calcon was carried out on a Nb/BDD anode at different current densities (7.5, 15 and 30 mA cm-2), while a stainless steel (SS) cathode was used for green H2 production. The results of the analysis by UV-vis spectroscopy and total organic carbon (TOC) clearly showed that the electrochemical oxidation (EO) of the Calcon dye after 180 min of electrolysis time by applying 30 mA cm-2 reached up to 90% of degradation and 57% of TOC removal. Meanwhile, under these experimental conditions, a green H2 production greater than 0.9 L was achieved, with a Faradaic efficiency of 98%. The hybrid electrolysis strategy is particularly attractive in the context of a circular economy, as these can be coupled with the use of more complex water matrices to transform organic depollution into an energy resource to produce H2 as a chemical energy carrier.

The impact of climate change on photovoltaic power potential in Southwestern Colombia.

In this paper, we present the first study of the long-term climate-change impact on photovoltaic power potential in Nariño, Colombia. In this region, more than half of the territory does not have a constant electricity supply, but it has great potential for solutions with renewable energy sources. Based on the Coordinated Regional Downscaling Experiment (CORDEX), we assess the change in photovoltaic power potential towards the end of this century, considering two climate change scenarios, one optimistic and the other pessimistic. Our results suggest that changes in photovoltaic power potential, by the end of the century, will have a maximum decrease of around 2.49% in the central zone of Nariño, with some non-affected areas, and a maximum increase of 2.52% on the southeastern side with respect to the pessimistic climate change scenario.

Instantaneous global horizontal irradiance and clearness index data with high temporal resolution for 1016 days.

Global Horizontal Irradiance was measured using a thermopile-type pyranometer during more than three years using a sample time of two seconds, with the purpose of capturing fast transient events of irradiance which are notable in tropical regions as the one where these data were collected: Bogotá, Colombia. The date and time of each measurement were registered along with the irradiance values. In addition, other related quantities were calculated and included for each one of the measurement instants: Optical Air Mass, Zenith angle, Extraterrestrial Solar Irradiance, and Clearness Index. Daily aggregated statistics of irradiance were calculated and are provided here too. Data points corresponding to nights were discarded. The raw data was analyzed to remove incomplete days, to guarantee that daily statistics are accurate and meaningful. After this data cleaning process, 1016 complete days remain, having a total of 21,959,912 data points. These data are useful for studying the effect of irradiance transients over photovoltaic systems, including power electronics, batteries and electric loads; it can also be used in studies about the stability of the radiative regime or the variability of irradiance such as Avila et al. (2019) (where part of these data was effectively used) and other related works cited there.

Design of a low-cost mobile multispectral albedometer with geopositioning and absolute orientation.

Albedo is the percentage of radiation that a given surface reflects. Its study is important to evaluate thermal effects in buildings, generation capacity with bifacial panels, among others. In this work, the design and validation of a low-cost mobile albedometer is presented, which measures the reflection in 8 spectral bands in the visible, additionally the system is equipped with a Global Navigation Satellite System (GNSS) receiver, to reference its position and an Inertial Measurement Unit (IMU) to know its absolute orientation, make corrections in real time or detect errors. The purpose of designing the mobile device is to measure a larger area and, since it is georeferenced, it is to feed GIS tools that allow designers to use the information.

Solar Cell Parameter Extraction Method from Illumination and Dark I-V Characteristics.

A novel method to extract the seven parameters of the double-diode model of solar cells using the current-voltage (I-V) characteristics under illumination and in the dark is presented. The algorithm consists of two subroutines which are alternatively run to adjust all the parameters of the cell in an iterative process. Curve fitting of the light I-V characteristics ensures accuracy in the prediction of the maximum power point, whereas simultaneously fitting the dark I-V characteristics results in a set of physically meaningful parameters that provide information about the physical performance of the photovoltaic devices. Experimental I-V curves of in-house solar cells are used to validate the proposed parameter extraction method, which can be furthermore applied to other types of p-n junction-based photovoltaic devices.

Effects on the emission of different interfaces formed between poly(3-methylthiophene) and poly(3-octylthiophene): electrochemical impedance and photoluminescence spectroscopy studies.

With the aim of verifying the optical properties of the systems formed by poly(3-methylthiophene) (P3MT) and poly(3-octylthiophene) (P3OT) on platinum (Pt) for use in organic photovoltaic device applications, electrochemical preparations of different interfaces with poly(3-alkylthiophenes) (P3ATs), synthesized both with 18 °C and without temperature control, were compared. These interfaces were prepared both as blends (Pt/P3MT:P3OT) and as layered films (Pt/P3MT/P3OT and Pt/P3OT/P3MT). Electrochemical impedance spectroscopy (EIS) was used to characterize the systems, and based on Bode-Phase diagrams, it was possible to monitor the stabilization of radical cation and dication segments of the thiophene ring. The findings corroborated previous studies by electrochemical spectroscopy and using in situ Raman spectroscopy under the same experimental conditions. We were able to verify the effects of experimental variables, such as synthesis temperature and different kinds of deposition. Temperature was found to be an extremely important factor in synthesis, since films synthesized at 18 °C favored the stabilization of radical cation segments in the polymer matrix, and layered deposition also favored the stabilization of these segments, since the layer closest to the electrode can act as an induction layer for the stability of radical cation segments in the system. Photoluminescence spectroscopy was used to verify the optical properties of the interfaces, in which occur the contributions of three segments in the P3ATs matrix. Thus, it has been demonstrated through photoluminescence decay time that the relative amount of radical cation and dication segments in the polymer matrix affects the lifetime of these segments in the different materials prepared, due to emission effects for these systems.

Solar Energy-Based Future Perspective for Organic Rankine Cycle Applications.

This article explores the patents of solar energy technologies in the organic Rankine cycle (ORC) applications. The conversion of low-quality thermal energy into electricity is one of the main characteristics of an ORC, making efficient and viable technologies available today. However, only a few and outdated articles that analyze patents that use solar energy technologies in ORC applications exist. This leads to a lack of updated information regarding the number of published patents, International Patent Classification (IPC) codes associated with them, technology life cycle status, and the most relevant patented developments. Thus, this article conducts a current investigation of patents published between January 2010 and May 2022 using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) methodology and keywords. One thousand two hundred ninety-nine patents were obtained as part of the study and classified in F and Y groups of the IPC. The time-lapse analyzed was between January 2010 and May 2022. In 2014 and 2015, a peak of published patents was observed. China (CN) was the country that published the most significant number of patents worldwide. However, the European Patent Office (EP), the World Intellectual Property Organization (WO), and the United States (US) publish the patents with the highest number of patent citations. Furthermore, the possible trend regarding the development of patents for each technology is presented. A high-performance theoretical ORC plant based on the patent information analyzed by this article is introduced. Finally, exploration of IPC revealed 17 codes related to solar energy technologies in ORC applications not indexed in the main search.

A low-cost system for real-time measuring of the sunlight incident angle using IoT.

The incidence angle of solar irradiance is an important parameter for sizing and locate photovoltaic systems, which affects the installation design and has a high influence in the power production of photovoltaic panels. This angle is traditionally estimated considering the geographical position, however, this approach ignores the existence of local elements that affect the generation, such as weather conditions, topography, constructions with high reflection, among others. Therefore, this work presents the design and construction of a measurement device with nine irradiance sensors, which are located at different angles on two orthogonal axes within a semisphere. Since the angles of the sensors are known, a model to determine the direction of the maximum incidence irradiance, at each instant of time, can be calculated from the on-site measurements. In this way, it is also possible to calculate the panel inclination and orientation producing the maximum power for a particular location. The device acquires the irradiance magnitude in the nine sensors in real time, and it is transmitted using the Internet to simplify data recollection. Finally, the device uses a low-cost platform, which makes possible the adoption of this solution in a wide range of applications, e.g. design, diagnostic or reconfiguration of PV arrays.