Weightless Neural Network-Based Detection and Diagnosis of Visual Faults in Photovoltaic Modules
<p>An overview of the proposed PV module data acquisition process.</p> "> Figure 2
<p>The complete methodology of the proposed WiSARD classifier.</p> "> Figure 3
<p>Feature selection process using DenseNet-201.</p> "> Figure 4
<p>(<b>a</b>) Simple difference among weightless and weighted neural networks. (<b>b</b>) A simplified view of RAM network and neuron (RAM devices).</p> "> Figure 5
<p>WiSARD classifier with discriminator module.</p> "> Figure 6
<p>Confusion matrix of WiSARD classifier with optimal hyperparameters.</p> ">
Abstract
:1. Introduction
- Weightless neural networks have not been applied in the diagnosis of PV module faults.
- Most of the published works utilized electroluminescence or thermal images in the detection of PV module faults, specifically, cracks and hotspots.
- To minimize the consumption of inspection duration, manpower and capital costs, unmanned aerial vehicles have been widely adopted.
- The lack of a public repository and acquisition of a PV module fault dataset is a challenge.
- Researchers have applied artificial intelligence-based techniques, namely, machine learning and deep learning, to fault diagnosis in PV modules.
- Over the years, deep learning or machine learning has been individually applied in the detection and diagnosis of PV module faults. However, a combined approach is still in the initial stages.
- First, there have been minimal attempts to use true color or RGB images for diagnosing faults in PV modules, guiding the direction of research. Since numerous works have studied the adoption of thermal or electroluminescence images.
- Secondly, UAVs for the acquisition of PV module images have been considered another driving force in reducing monitoring time and manpower.
- Further, a combined approach of machine and deep learning in the process of fault detection can be considered a prime innovation.
- Finally, to the best of the authors’ modest knowledge, this is the first study where a WNN is applied to diagnose and classify visual faults in PV modules.
- A PV module dataset was created with six different conditions like good panel, delamination, snail trail, burn marks, discoloration and glass breakage. The dataset was acquired using an unmanned aerial vehicle and pre-processed for further usage.
- The obtained dataset was augmented artificially through application image transforms to enhance the learning capability of the pre-trained network adopted. The augmented dataset was split into training and testing datasets (with a ratio of 0.8:0.2) resized to a size of 224 × 224 to be fed as input to DenseNet-201.
- The features from the final fully connected layer (fc1000) of DenseNet-201 were extracted and saved as a data file. These extracted features were passed onto a J48 decision tree to select the impactful and contributing features.
- The selected features were used to train the WNN (WiSARD classifier). The trained model was used to classify the PV module conditions with the supplied test dataset.
- To improve the classification accuracy of the adopted WNN, several hyperparameters like tic number, bleach configuration, bit number, bleach flag, map type and bleach step were configured.
2. Experimental Studies
3. Methodology
3.1. DenseNet-201-Based Feature Extraction
3.2. Selection of Features
3.3. WNN-Based Classification (WiSARD Classifier)
- Step 1—Selection of attributes
- Step 2—Input pattern mapping through simple mapping, weighted moving average by hamming distance, weighted moving average by degree of membership, weighted moving average by simple exponential smoothing
- Step 3—Structural determination of neural networks in the WiSARD classifier
- Step 4—Combining or clustering of filters
- Step 5—Training and testing of the WiSARD classifier
4. Results and Discussion
4.1. Impact of Changing the ‘Bit Number’
4.2. Impact of Changing Bleach Confidence
4.3. Impact of Changing the ‘Bleach Flag’
4.4. Impact of Changing the ‘Bleach Step’
4.5. Impact of the ‘Map Type’ Hyperparameter
- Linear map: The input characteristics are spread and sequentially related to the bits in the linear map type. Each feature is assigned to a different bit, resulting in a linear mapping from the input pattern to the bit cells. This means that the first feature is linked to the first bit, the second to the second bit and so on. The linear map offers a simple and predictable mapping pattern.
- Random map: The input characteristics are dispersed and linked to the bits at random in the random map type. The mapping from the input pattern to the bit cells is non-sequential, i.e., random. Irrespective of the order, any feature can link to any of the bits in the bit cells. The random map adds more variety and randomization to the mapping process.
4.6. Impact of the ‘Tic Number’ Hyperparameter
4.7. Optimal Hyperparameter Selection
4.8. Comparison with Other Studies
5. Conclusions
Supplementary Materials
Author Contributions
Funding
Data Availability Statement
Acknowledgments
Conflicts of Interest
References
- IRENA. World Energy Transitions Outlook: 1.5 °C Pathway; IRENA: Masdar City, United Arab Emirates, 2022. [Google Scholar]
- Bahar, N.H.A.; Lo, M.; Sanjaya, M.; Van Vianen, J.; Alexander, P.; Ickowitz, A.; Sunderland, T. Meeting the food security challenge for nine billion people in 2050: What impact on forests? Glob. Environ. Chang. 2020, 62, 102056. [Google Scholar]
- Li, Z.; Luan, R.; Lin, B. The trend and factors affecting renewable energy distribution and disparity across countries. Energy 2022, 254, 124265. [Google Scholar] [CrossRef]
- Yoro, K.O.; Daramola, M.O. CO2 emission sources, greenhouse gases, and the global warming effect. In Advances in Carbon Capture: Methods, Technologies and Applications; Elsevier: Amsterdam, The Netherlands, 2020. [Google Scholar]
- Hassan, R.; Das, B.K.; Hasan, M. Integrated off-grid hybrid renewable energy system optimization based on economic, environmental, and social indicators for sustainable development. Energy 2022, 250, 123823. [Google Scholar] [CrossRef]
- Kellil, N.; Aissat, A.; Mellit, A. Fault diagnosis of photovoltaic modules using deep neural networks and infrared images under Algerian climatic conditions. Energy 2023, 263, 125902. [Google Scholar] [CrossRef]
- Mellit, A.; Tina, G.M.; Kalogirou, S.A. Fault detection and diagnosis methods for photovoltaic systems: A review. Renew. Sustain. Energy Rev. 2018, 91, 1–17. [Google Scholar] [CrossRef]
- Wang, H.; Zhao, J.; Sun, Q.; Zhu, H. Probability modeling for PV array output interval and its application in fault diagnosis. Energy 2019, 189, 116248. [Google Scholar] [CrossRef]
- Aboagye, B.; Gyamfi, S.; Ofosu, E.A.; Djordjevic, S. Degradation analysis of installed solar photovoltaic (PV) modules under outdoor conditions in Ghana. Energy Rep. 2021, 7, 6921–6931. [Google Scholar] [CrossRef]
- Chouder, A.; Silvestre, S. Automatic supervision and fault detection of PV systems based on power losses analysis. Energy Convers. Manag. 2010, 51, 1929–1937. [Google Scholar] [CrossRef]
- Gallardo-Saavedra, S.; Hernández-Callejo, L.; Alonso-García, M.d.C.; Santos, J.D.; Morales-Aragonés, J.I.; Alonso-Gómez, V.; Moreton, A.; Gonzalez, M.A.; Martinez-Sacristan, O. Nondestructive characterization of solar PV cells defects by means of electroluminescence, infrared thermography, I–V curves and visual tests: Experimental study and comparison. Energy 2020, 205, 117930. [Google Scholar] [CrossRef]
- Dhimish, M.; Holmes, V.; Mehrdadi, B.; Dales, M. The impact of cracks on photovoltaic power performance. J. Sci. Adv. Mater. Devices 2017, 2, 199–209. [Google Scholar] [CrossRef]
- Hu, Y.; Zhang, J.; Wu, J.; Cao, W.; Tian, G.Y.; Kirtley, J.L. Efficiency Improvement of Nonuniformly Aged PV Arrays. IEEE Trans. Power Electron. 2016, 32, 1124–1137. [Google Scholar] [CrossRef] [Green Version]
- Akram, M.W.; Li, G.; Jin, Y.; Chen, X.; Zhu, C.; Zhao, X.; Khaliq, A.; Faheem, M.; Ahmad, A. CNN based automatic detection of photovoltaic cell defects in electroluminescence images. Energy 2019, 189, 116319. [Google Scholar] [CrossRef]
- Chen, H.; Zhao, H.; Han, D.; Liu, K. Accurate and robust crack detection using steerable evidence filtering in electroluminescence images of solar cells. Opt. Lasers Eng. 2019, 118, 22–33. [Google Scholar] [CrossRef]
- Sridharan, N.V.; Sugumaran, V. Convolutional Neural Network based Automatic Detection of Visible Faults in a Photovoltaic Module. Energy Sources Part A Recovery Util. Environ. Eff. 2021, 1–16. [Google Scholar] [CrossRef]
- S, N.V.; Sugumaran, V. Fault diagnosis of visual faults in photovoltaic modules: A Review. Int. J. Green Energy 2020, 18, 37–50. [Google Scholar] [CrossRef]
- Lindahl, J.; Lingfors, D.; Elmqvist, Å.; Mignon, I. Economic analysis of the early market of centralized photovoltaic parks in Sweden. Renew. Energy 2022, 185, 1192–1208. [Google Scholar] [CrossRef]
- Ullah, N.; Sami, I.; Jamal Babqi, A.; Alkhammash, I.H.; Belkhier, Y.; Althobaiti, A. Processor in the Loop Verification of Fault Tolerant Control for a Three Phase Inverter in Grid Connected PV System. Energy Sources Part A Recovery Util. Environ. Eff. 2023, 45, 3760–3776. [Google Scholar] [CrossRef]
- Fonseca Alves, R.H.; Deus Júnior, G.A.d.; Marra, E.G.; Lemos, R.P. Automatic fault classification in photovoltaic modules using Convolutional Neural Networks. Renew. Energy 2021, 179, 502–516. [Google Scholar] [CrossRef]
- Otamendi, U.; Martinez, I.; Quartulli, M.; Olaizola, I.G.; Viles, E.; Cambarau, W. Segmentation of cell-level anomalies in electroluminescence images of photovoltaic modules. Sol. Energy 2021, 220, 914–926. [Google Scholar] [CrossRef]
- Mellit, A.; Kalogirou, S. Artificial intelligence and internet of things to improve efficacy of diagnosis and remote sensing of solar photovoltaic systems: Challenges, recommendations and future directions. Renew. Sustain. Energy Rev. 2021, 143, 110889. [Google Scholar]
- Sridharan, N.V.; Sugumaran, V. Visual fault detection in photovoltaic modules using decision tree algorithms with deep learning features. Energy Sources Part A Recovery Util. Environ. Eff. 2021, 1–17. [Google Scholar] [CrossRef]
- Sizkouhi, A.M.; Aghaei, M.; Esmailifar, S.M. A deep convolutional encoder-decoder architecture for autonomous fault detection of PV plants using multi-copters. Sol. Energy 2021, 223, 217–228. [Google Scholar] [CrossRef]
- Tsanakas, J.A.; Botsaris, P.N. Passive and Active Thermographic Assessment as a Tool for Condition-Based Performance Monitoring of Photovoltaic Modules. J. Sol. Energy Eng. 2011, 133, 021012. [Google Scholar] [CrossRef]
- Tsanakas, J.A.; Botsaris, P.N. An infrared thermographic approach as a hot-spot detection tool for photovoltaic modules using image histogram and line profile analysis. Int. J. Cond. Monit. 2012, 2, 22–30. [Google Scholar] [CrossRef]
- Tsanakas, J.A.; Ha, L.; Buerhop, C. Faults and infrared thermographic diagnosis in operating c-Si photovoltaic modules: A review of research and future challenges. Renew. Sustain. Energy Rev. 2016, 62, 695–709. [Google Scholar] [CrossRef]
- Baig, H.; Jani, R.; Markam, B.K.; Maiti, S.; Mallick, T.K. Modelling and experimental analysis of a seasonally tracked V-trough PV/T system in India. Sol. Energy 2018, 170, 618–632. [Google Scholar] [CrossRef]
- Eskandari, A.; Milimonfared, J.; Aghaei, M. Line-line fault detection and classification for photovoltaic systems using ensemble learning model based on I-V characteristics. Sol. Energy 2020, 211, 354–365. [Google Scholar] [CrossRef]
- Grimaccia, F.; Leva, S.; Dolara, A.; Aghaei, M. Survey on PV Modules’ Common Faults After an O&M Flight Extensive Campaign Over Different Plants in Italy. IEEE J. Photovolt. 2017, 7, 810–816. [Google Scholar] [CrossRef] [Green Version]
- Grimaccia, F.; Leva, S.; Niccolai, A.; Cantoro, G. Assessment of PV Plant Monitoring System by Means of Unmanned Aerial Vehicles. In Proceedings of the 2018 IEEE International Conference on Environment and Electrical Engineering and 2018 IEEE Industrial and Commercial Power Systems Europe, EEEIC/I and CPS Europe 2018, Palermo, Italy, 12–15 June 2018. [Google Scholar]
- Tsanakas, J.A.; Ha, L.D.; Al Shakarchi, F. Advanced inspection of photovoltaic installations by aerial triangulation and terrestrial georeferencing of thermal/visual imagery. Renew. Energy 2017, 102, 224–233. [Google Scholar] [CrossRef]
- Leva, S.; Aghaei, M.; Grimaccia, F. PV power plant inspection by UAS: Correlation between altitude and detection of defects on PV modules. In Proceedings of the 2015 IEEE 15th International Conference on Environment and Electrical Engineering, EEEIC 2015—Conference Proceedings, Rome, Italy, 10–13 June 2015. [Google Scholar]
- Tsanakas, J.; Chrysostomou, D.; Botsaris, P.; Gasteratos, A. Fault diagnosis of photovoltaic modules through image processing and Canny edge detection on field thermographic measurements. Int. J. Sustain. Energy 2013, 34, 351–372. [Google Scholar] [CrossRef]
- Li, X.; Yang, Q.; Chen, Z.; Luo, X.; Yan, W. Visible defects detection based on UAV-based inspection in large-scale photovoltaic systems. IET Renew. Power Gener. 2017, 11, 1234–1244. [Google Scholar] [CrossRef]
- Dashtdar, M.; Sarada, K.; Hosseinimoghadam, S.M.S.; Kalyan, C.H.N.S.; Venkateswarlu, A.N.; Goud, B.S.; Reddy, C.H.R.; Belkhier, Y.; Bajaj, M.; Reddy, B.N. Faulted Section Identification and Fault Location in Power Network Based on Histogram Analysis of Three-phase Current and Voltage Modulated. J. Electr. Eng. Technol. 2022, 17, 2631–2647. [Google Scholar] [CrossRef]
- Venkatesh, S.N.; Sugumaran, V. A combined approach of convolutional neural networks and machine learning for visual fault classification in photovoltaic modules. Proc. Inst. Mech. Eng. Part O J. Risk Reliab. 2021, 236, 148–159. [Google Scholar] [CrossRef]
- Muturatnam, A.B.; Sridharan, N.V.; Sreelatha, A.P.; Vaithiyanathan, S. Enhanced Tyre Pressure Monitoring System for Nitrogen Filled Tyres Using Deep Learning. Machines 2023, 11, 434. [Google Scholar] [CrossRef]
- Venkatesh, S.N.; Balaji, P.A.; Elangovan, M.; Annamalai, K.; Indira, V.; Sugumaran, V.; Mahamuni, V.S. Transfer Learning-Based Condition Monitoring of Single Point Cutting Tool. Comput. Intell. Neurosci. 2022, 2022, 3205960. [Google Scholar] [CrossRef]
- Venkatesh, S.N.; Chakrapani, G.; Senapti, S.B.; Annamalai, K.; Elangovan, M.; Indira, V.; Sugumaran, V.; Mahamuni, V.S. Misfire Detection in Spark Ignition Engine Using Transfer Learning. Comput. Intell. Neurosci. 2022, 2022, 7606896. [Google Scholar] [CrossRef]
- Venkatesh, S.N.; Sugumaran, V. Fault Detection in aerial images of photovoltaic modules based on Deep learning. IOP Conf. Ser. Mater. Sci. Eng. 2021, 1012, 012030. [Google Scholar] [CrossRef]
- Chouay, Y.; Ouassaid, M. A Multi-stage SVM Based Diagnosis Technique for Photovoltaic PV Systems. In Advances in Intelligent Systems and Computing, 1350 AISC; Springer: Berlin/Heidelberg, Germany, 2021; pp. 183–193. [Google Scholar]
- Niazi, K.A.K.; Akhtar, W.; Khan, H.A.; Yang, Y.; Athar, S. Hotspot diagnosis for solar photovoltaic modules using a Naive Bayes classifier. Sol. Energy 2019, 190, 34–43. [Google Scholar] [CrossRef]
- Naveen Venkatesh, S.; Sugumaran, V. Machine vision based fault diagnosis of photovoltaic modules using lazy learning approach. Meas. J. Int. Meas. Confed. 2022, 191, 110786. [Google Scholar] [CrossRef]
- Aleksander, I. Adaptive systems of logic networks and binary memories. In Proceedings of the AFIPS Conference Proceedings—1967 Spring Joint Computer Conference, AFIPS 1967, Atlantic, NJ, USA, 18–20 April 1967. [Google Scholar]
- Lei, Y.; He, Z.; Zi, Y. EEMD method and WNN for fault diagnosis of locomotive roller bearings. Expert Syst. Appl. 2011, 38, 7334–7341. [Google Scholar] [CrossRef]
- De Souza, A.F.; Freitas, F.D.; De Almeida, A.G.C. High performance prediction of stock returns with VG-RAM weightless neural networks. In Proceedings of the 3rd Workshop on High Performance Computational Finance, WHPCF 2010, New Orleans, LA, USA, 14 November 2010. [Google Scholar]
- Cardoso, D.O.; Lima, P.M.V.; de Gregorio, M.; Gama, J.; França, F.M.G. Clustering data streams with weightless neural networks. In Proceedings of the ESANN 2011—19th European Symposium on Artificial Neural Networks, Bruges, Belgium, 27–29 April 2011. [Google Scholar]
- McElroy, B.; Howells, G. Automated adaptation of input and output data for a weightless artificial neural network. Int. J. Database Theory Appl. 2011, 4, 37–46. [Google Scholar]
- Subhashini, R.; Nagarajan, E.; Niveditha, P.R. Detection of an incognitos intruder in industries and semantic mapping of emotions. Int. J. Appl. Eng. Res. 2014, 9, 6727–6734. [Google Scholar]
- Grieco, B.P.A.; Lima, P.M.V.; De Gregorio, M.; França, F.M.G. Producing pattern examples from “mental” images. Neurocomputing 2010, 73, 1057–1064. [Google Scholar] [CrossRef]
- De Gregorio, M.; Giordano, M. Change detection with weightless neural networks. In Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, Columbus, OH, USA, 23–28 June 2014. [Google Scholar]
- Cardoso, D.; De Gregorio, M.; Lima, P.; Gama, J.; França, F. A weightless neural network-based approach for stream data clustering. In Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Springer: Berlin/Heidelberg, Germany, 2012. [Google Scholar]
- Köntges, M.; Kurtz, S.; Packard, C.E.; Jahn, U.; Berger, K.; Kato, K.; Friesen, T.; Liu, H.; Van Iseghem, M. Review of Failures of Photovoltaic Modules; IEA-PVPS T13-01:2014; IEA: Paris, France, 2014. [Google Scholar]
- Chandel, S.; Naik, M.N.; Sharma, V.; Chandel, R. Degradation analysis of 28 year field exposed mono-c-Si photovoltaic modules of a direct coupled solar water pumping system in western Himalayan region of India. Renew. Energy 2015, 78, 193–202. [Google Scholar] [CrossRef]
- Dolara, A.; Lazaroiu, G.C.; Leva, S.; Manzolini, G.; Votta, L. Snail Trails and Cell Microcrack Impact on PV Module Maximum Power and Energy Production. IEEE J. Photovolt. 2016, 6, 1269–1277. [Google Scholar] [CrossRef] [Green Version]
- Sánchez-Friera, P.; Piliougine, M.; Peláez, J.; Carretero, J.; De Cardona, M.S. Analysis of degradation mechanisms of crystalline silicon PV modules after 12 years of operation in Southern Europe. Prog. Photovolt. Res. Appl. 2011, 19, 658–666. [Google Scholar] [CrossRef]
- Jeong, J. Microelectronics Reliability Effect of EVA discoloration in 25-year-old single crystalline silicon photovoltaic modules operated under moderate climate. Microelectron. Reliab. 2022, 138, 114721. [Google Scholar] [CrossRef]
- Mehta, S.; Azad, A.P.; Chemmengath, S.A.; Raykar, V.; Kalyanaraman, S. DeepSolarEye: Power Loss Prediction and Weakly Supervised Soiling Localization via Fully Convolutional Networks for Solar Panels. In Proceedings of the 2018 IEEE Winter Conference on Applications of Computer Vision, WACV 2018, Lake Tahoe, NV, USA, 12–15 March 2018; pp. 333–342. [Google Scholar]
- Wang, J.; Zhao, B.; Yao, X. PV Abnormal Shading Detection Based on Convolutional Neural Network. In Proceedings of the 32nd Chinese Control and Decision Conference, Hefei, China, 22–24 August 2020; pp. 1580–1583. [Google Scholar] [CrossRef]
- Li, X.; Yang, Q.; Wang, J.; Chen, Z.; Yan, W. Intelligent fault pattern recognition of aerial photovoltaic module images based on deep learning technique. In Proceedings of the IMCIC 2018—9th International Multi-Conference on Complexity, Informatics and Cybernetics, Proceedings 2018, Orlando, FL, USA, 13–16 March 2018; Volume 1, pp. 22–27. [Google Scholar]
- Chen, H.; Pang, Y.; Hu, Q.; Liu, K. Solar cell surface defect inspection based on multispectral convolutional neural network. J. Intell. Manuf. 2018, 31, 453–468. [Google Scholar] [CrossRef] [Green Version]
- Li, X.; Yang, Q.; Lou, Z.; Yan, W. Deep Learning Based Module Defect Analysis for Large-Scale Photovoltaic Farms. IEEE Trans. Energy Convers. 2018, 34, 520–529. [Google Scholar] [CrossRef]
- Sridharan, N.V.; Sugumaran, V. Deep learning-based ensemble model for classification of photovoltaic module visual faults. Energy Sources Part A Recovery Util. Environ. Eff. 2022, 44, 5287–5302. [Google Scholar] [CrossRef]
Model Name | Efficiency | Number of Cells | Type | Weight | Current | Maximum Power | Maximum Power Point Current (Impp) | Dimensions |
---|---|---|---|---|---|---|---|---|
USP-36 | 9–10% | 36 | Monocrystalline | 3.5 kg | 2.25 A | 36 W | 2.1 | (1011 × 435 × 36) mm |
Transformation Operation | Noise | Angle of Rotation | Blur | Flip Angle | Warp |
---|---|---|---|---|---|
Value | Random | 0–180° | Gaussian | 90° | 40 |
S. No. | Visual Faults in PVM | Reason for Occurrence of Fault | Effect on Modules | Images |
---|---|---|---|---|
1 | Burn marks [54] | Solder bond failure, ribbon ripping, localized heating. | Safety risks and a drop in performance. | |
2 | Glass breakage [55] | Thermal stresses, physical injury induced during installation and travel. | Lower radiance, corrosion and moisture invasion. | |
3 | Snail trail [56] | Microcracks around borders caused by stress. | Quicker degradation. | |
4 | Delamination [57] | Adhesion failure between the glass, encapsulant and back cover. | Moisture ingresses caused by corrosion. | |
5 | Discoloration [58] | Increased heat, humidity and UV radiation exposure. | Power loss, physical color changes in modules (yellowing or browning). |
Bit Number | Training Set Accuracy (%) | Cross Validation Accuracy (%) | Test Set Accuracy (%) |
---|---|---|---|
4 | 100.00 | 99.28 | 99.68 |
8 | 100.00 | 99.48 | 100.00 |
16 | 100.00 | 99.80 | 100.00 |
32 | 100.00 | 99.92 | 100.00 |
Bleach Confidence | Training Set Accuracy (%) | Cross Validation Accuracy (%) | Test Set Accuracy (%) |
---|---|---|---|
0.60 | 100.00 | 99.74 | 100.00 |
0.70 | 100.00 | 99.70 | 99.84 |
0.80 | 100.00 | 99.76 | 100.00 |
0.90 | 100.00 | 99.80 | 100.00 |
1.00 | 100.00 | 99.84 | 100.00 |
Bleach Flag | Training Set Accuracy (%) | Cross Validation Accuracy (%) | Test Set Accuracy (%) |
---|---|---|---|
TRUE | 58.33 | 46.54 | 40.47 |
FALSE | 100.00 | 99.84 | 100.00 |
Bleach Step | Training Set Accuracy (%) | Cross Validation Accuracy (%) | Test Set Accuracy (%) |
---|---|---|---|
1 | 100.00 | 99.88 | 100.00 |
2 | 100.00 | 99.76 | 100.00 |
5 | 100.00 | 99.76 | 100.00 |
10 | 100.00 | 99.84 | 100.00 |
Map Type | Training Set Accuracy (%) | Cross Validation Accuracy (%) | Test Set Accuracy (%) |
---|---|---|---|
RANDOM | 100.00 | 99.88 | 100.00 |
LINEAR | 100.00 | 90.91 | 91.58 |
Tic Number | Training Set Accuracy (%) | Cross Validation Accuracy (%) | Test Set Accuracy (%) |
---|---|---|---|
1 | 20.23 | 20.00 | 20.30 |
10 | 100.00 | 98.21 | 97.46 |
20 | 100.00 | 98.84 | 99.68 |
50 | 100.00 | 99.28 | 99.84 |
100 | 100.00 | 99.72 | 99.84 |
256 | 100.00 | 99.88 | 100.00 |
Hyperparameter | Configuration |
---|---|
bits | 32 |
bleach confidence | 1.00 |
bleach flag | FALSE |
bleach step | 1 |
map type | RANDOM |
tic number | 256 |
Reference | Methodology Used | Classification Accuracy (%) | Test Time (s) |
---|---|---|---|
[59] | DeepSolarEye (ResNet-Based) | 97.80 | - |
[60] | Yolo V3 | 96.30 | - |
[61] | Custom CNN | 79.06 | - |
[62] | Custom CNN | 94.30 | 3.66 |
[63] | Custom CNN | 97.90 | 0.55 |
[16] | Custom CNN | 95.07 | - |
[23] | Pretrained CNN + Random Forest | 98.25 | 0.89 |
[44] | Pretrained CNN + K-Nearest Neighbor | 98.95 | 0.04 |
[64] | Ensemble Model | 99.04 | 2.50 |
Proposed | DenseNet-201+ WiSARD | 100.00 | 1.44 |
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Sridharan, N.V.; Joseph, J.V.; Vaithiyanathan, S.; Aghaei, M. Weightless Neural Network-Based Detection and Diagnosis of Visual Faults in Photovoltaic Modules. Energies 2023, 16, 5824. https://doi.org/10.3390/en16155824
Sridharan NV, Joseph JV, Vaithiyanathan S, Aghaei M. Weightless Neural Network-Based Detection and Diagnosis of Visual Faults in Photovoltaic Modules. Energies. 2023; 16(15):5824. https://doi.org/10.3390/en16155824
Chicago/Turabian StyleSridharan, Naveen Venkatesh, Jerome Vasanth Joseph, Sugumaran Vaithiyanathan, and Mohammadreza Aghaei. 2023. "Weightless Neural Network-Based Detection and Diagnosis of Visual Faults in Photovoltaic Modules" Energies 16, no. 15: 5824. https://doi.org/10.3390/en16155824