{"title":"Modeling and Detection of PV Panel Hard Shading Using Artificial Neural Network","authors":"Bryan E. Escoto","doi":"10.1109/CONIT55038.2022.9847671","DOIUrl":null,"url":null,"abstract":"The hard shading caused by dirt accumulation on the PV surface, shadows caused by near structures, trees, and even unwanted materials at the surface of the panels degrades the performance of the PV panels, significantly reducing the power output and its efficiency. However, the detection of shading of the solar panel is a complex and challenging process since the panel's power conversion varies and is affected by several factors such as solar irradiance, temperature, the position of the sun, location of shading, etc. This project created an Artificial Neural Network (ANN) model that detects the hard shading and its coverage to PV panels. The optimum ANN model developed in this study can detect solar panel hard shading coverage with 99.98 % accuracy. The best ANN network topology is 3-60-1 (input-hidden neurons-output) model which provides an excellent generalization ability. This model utilized the tan Sigmoid transfer function for both input-hidden and hidden-output layer, and for the optimization process, Levenberg Marquardt outperformed other algorithms. The optimum ANN model has the lowest MSE value of 0.000020333 and with highest R-values of 0.99992, 0.99989, 0.9999, and 0.99992 for training, validation, testing, and overall, respectively. Based on the sensitivity analysis result, the open-circuit voltage significantly contributes to the solar panel shading detection.","PeriodicalId":270445,"journal":{"name":"2022 2nd International Conference on Intelligent Technologies (CONIT)","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2022-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"1","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2022 2nd International Conference on Intelligent Technologies (CONIT)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/CONIT55038.2022.9847671","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 1
Abstract
The hard shading caused by dirt accumulation on the PV surface, shadows caused by near structures, trees, and even unwanted materials at the surface of the panels degrades the performance of the PV panels, significantly reducing the power output and its efficiency. However, the detection of shading of the solar panel is a complex and challenging process since the panel's power conversion varies and is affected by several factors such as solar irradiance, temperature, the position of the sun, location of shading, etc. This project created an Artificial Neural Network (ANN) model that detects the hard shading and its coverage to PV panels. The optimum ANN model developed in this study can detect solar panel hard shading coverage with 99.98 % accuracy. The best ANN network topology is 3-60-1 (input-hidden neurons-output) model which provides an excellent generalization ability. This model utilized the tan Sigmoid transfer function for both input-hidden and hidden-output layer, and for the optimization process, Levenberg Marquardt outperformed other algorithms. The optimum ANN model has the lowest MSE value of 0.000020333 and with highest R-values of 0.99992, 0.99989, 0.9999, and 0.99992 for training, validation, testing, and overall, respectively. Based on the sensitivity analysis result, the open-circuit voltage significantly contributes to the solar panel shading detection.