为了减轻在光伏板表面缺陷检测中由图像噪声和目标过小造成的影响,提出了改进Dn-YOLOv7(de-noising you only look once version 7)算法。该算法结合去噪卷积神经网络(de-noising convolutional neural network,DnCNN),提出了一个降噪模...为了减轻在光伏板表面缺陷检测中由图像噪声和目标过小造成的影响,提出了改进Dn-YOLOv7(de-noising you only look once version 7)算法。该算法结合去噪卷积神经网络(de-noising convolutional neural network,DnCNN),提出了一个降噪模块(de-noise block,DnBlock),该模块使用了噪声容限更大的损失函数,并利用空间坐标卷积(coordinates convolution,CoordConv)对噪声通道进行卷积整合,增强了网络的降噪能力。同时,使用归一化高斯瓦瑟斯坦距离(normalized Gaussian Wasserstein distance,NWD)替代传统交并比(intersection over union,IoU)损失函数,提高了模型对小目标的检测能力。结果表明,改进后的模型具有降噪能力并提高了检测精确率,在无噪声水平下平均精确率均值达到了96.6%,在较强的高斯噪声和脉冲噪声下平均精确率均值分别达到91.4%和85.4%,检测速度达到78.0帧/s。该算法在航拍图像的光伏板缺陷检测中有一定实用价值。展开更多
Organic photovoltaic(OPV) devices hold great promise for indoor light harvesting,offering a theoretical upper limit of power conversion efficiency that surpasses that of other photovoltaic technologies.However,the pre...Organic photovoltaic(OPV) devices hold great promise for indoor light harvesting,offering a theoretical upper limit of power conversion efficiency that surpasses that of other photovoltaic technologies.However,the presence of high leakage currents in OPV devices commonly constrains their effective performance under indoor conditions.In this study,we identified that the origin of the high leakage currents in OPV devices lay in pinhole defects present within the active layer(AL).By integrating an automated spin-coating strategy with sequential deposition processes,we achieved the compactness of the AL and minimized the occurrence of pinhole defects therein.Experimental findings demonstrated that with an increase in the number of deposition cycles,the density of pinhole defects in the AL underwent a marked reduction.Consequently,the leakage current experienced a substantial decrease by several orders of magnitude which achieved through well-calibrated AL deposition procedures.This enabled a twofold enhancement in the power conversion efficiency(PCE) of the OPV devices under conditions of indoor illumination.展开更多
This paper proposes a voltage-based hot-spot detection method for defective cells in PV module using projector. The presence of internal crystal defects is one of the main causes of hot-spot phenomenon in PV modules. ...This paper proposes a voltage-based hot-spot detection method for defective cells in PV module using projector. The presence of internal crystal defects is one of the main causes of hot-spot phenomenon in PV modules. Authors previously investigated the physical characteristics of hot-spot phenomenon referring to internal crystal defect. Based on it, a hot-spot detection method named as current-based SRC （self reverse current） detection method is developed. However, it becomes extraordinarily complicated to determine the defective cells under low illumination. In order to avoid this disadvantage, authors improve the SRC detection method by applying voltage. From the feasibility experiment results, it is confirmed that by calculating cell HSI （hotspots index） with voltage, the PV modules with defective cells can be prospectively excluded even under low illumination.展开更多
Chalcogenide perovskites(CPs) exhibiting lower band gaps than oxide perovskites and higher stability than halide perovskites are promising materials for photovoltaic and optoelectronic applications. For such applicati...Chalcogenide perovskites(CPs) exhibiting lower band gaps than oxide perovskites and higher stability than halide perovskites are promising materials for photovoltaic and optoelectronic applications. For such applications, the absence of deep defect levels serving as recombination centers(dubbed defect tolerance) is a highly desirable property. Here,using density functional theory(DFT) calculations, we study the intrinsic defects in BaZrS_(3), a representative CP material.We compare Hubbard-U and hybrid functional methods, both of which have been widely used in addressing the band gap problem of semi-local functionals in DFT. We find that tuning the U value to obtain experimental bulk band gap and then using the obtained U value for defect calculations may result in over-localization of defect states. In the hybrid functional calculation, the band gap of BaZrS_(3)can be accurately obtained. We observe the formation of small S-atom clusters in both methods, which tend to self-passivate the defects from forming mid-gap levels. Even though in the hybrid functional calculations several relatively deep defects are observed, all of them exhibit too high formation energy to play a significant role if the materials are prepared under thermal equilibrium.BaZrS_(3)is thus expected to exhibit sufficient defect tolerance promising for photovoltaic and optoelectronic applications.展开更多
Tin perovskite solar cells(TPSCs)are promising for lead-free perovskite solar cells(PSCs)and have led to extensive research;however,the poor crystallinity and chemical stability of tin perovskites are two issues that ...Tin perovskite solar cells(TPSCs)are promising for lead-free perovskite solar cells(PSCs)and have led to extensive research;however,the poor crystallinity and chemical stability of tin perovskites are two issues that prevent stable TPSCs.In this study,we outline a new process that addresses these issues by using tin(II)acetate(Sn(Ac)2)in place of the conventional SnF2 precursor additive.Compared with SnF2,Sn(Ac)2 improves the crystallinity and stability of tin perovskite with fewer defects and better charge extraction.Using this process,we developed a device that has a higher external quantum efficiency for charge extraction compared with the control devices and a power conversion efficiency of 9.93%,which maintained more than 90%of its initial efficiency after 1000 h operation at the maximum power point under standard AM 1.5G solar illumination.展开更多
文摘为了减轻在光伏板表面缺陷检测中由图像噪声和目标过小造成的影响,提出了改进Dn-YOLOv7(de-noising you only look once version 7)算法。该算法结合去噪卷积神经网络(de-noising convolutional neural network,DnCNN),提出了一个降噪模块(de-noise block,DnBlock),该模块使用了噪声容限更大的损失函数,并利用空间坐标卷积(coordinates convolution,CoordConv)对噪声通道进行卷积整合,增强了网络的降噪能力。同时,使用归一化高斯瓦瑟斯坦距离(normalized Gaussian Wasserstein distance,NWD)替代传统交并比(intersection over union,IoU)损失函数,提高了模型对小目标的检测能力。结果表明,改进后的模型具有降噪能力并提高了检测精确率,在无噪声水平下平均精确率均值达到了96.6%,在较强的高斯噪声和脉冲噪声下平均精确率均值分别达到91.4%和85.4%,检测速度达到78.0帧/s。该算法在航拍图像的光伏板缺陷检测中有一定实用价值。
基金Fundamental Research Funds for the Central Universities,China (No. 2232022A13)。
文摘Organic photovoltaic(OPV) devices hold great promise for indoor light harvesting,offering a theoretical upper limit of power conversion efficiency that surpasses that of other photovoltaic technologies.However,the presence of high leakage currents in OPV devices commonly constrains their effective performance under indoor conditions.In this study,we identified that the origin of the high leakage currents in OPV devices lay in pinhole defects present within the active layer(AL).By integrating an automated spin-coating strategy with sequential deposition processes,we achieved the compactness of the AL and minimized the occurrence of pinhole defects therein.Experimental findings demonstrated that with an increase in the number of deposition cycles,the density of pinhole defects in the AL underwent a marked reduction.Consequently,the leakage current experienced a substantial decrease by several orders of magnitude which achieved through well-calibrated AL deposition procedures.This enabled a twofold enhancement in the power conversion efficiency(PCE) of the OPV devices under conditions of indoor illumination.
文摘This paper proposes a voltage-based hot-spot detection method for defective cells in PV module using projector. The presence of internal crystal defects is one of the main causes of hot-spot phenomenon in PV modules. Authors previously investigated the physical characteristics of hot-spot phenomenon referring to internal crystal defect. Based on it, a hot-spot detection method named as current-based SRC （self reverse current） detection method is developed. However, it becomes extraordinarily complicated to determine the defective cells under low illumination. In order to avoid this disadvantage, authors improve the SRC detection method by applying voltage. From the feasibility experiment results, it is confirmed that by calculating cell HSI （hotspots index） with voltage, the PV modules with defective cells can be prospectively excluded even under low illumination.
基金supported by the National Natural Science Foundation of China (11774365)the Natural Science Foundation of Shanghai (19ZR1421800)+4 种基金Shanghai International Cooperation Project (20520760900)the Opening Project and Science Foundation for Youth Scholar of State Key Laboratory of High Performance Ceramics and Superfine Microstructures (SKL201804 and SKL201803SIC) support by US National Science Foundation (NSF) (CBET1510121)US Department of Energy (DOE) (DEEE0007364)support by US NSF (CBET-1510948).support by US NSF (DMR-1506669)support by the Fundamental Research Funds for the Central Universities (DUT21RC(3) 033)。
文摘Chalcogenide perovskites(CPs) exhibiting lower band gaps than oxide perovskites and higher stability than halide perovskites are promising materials for photovoltaic and optoelectronic applications. For such applications, the absence of deep defect levels serving as recombination centers(dubbed defect tolerance) is a highly desirable property. Here,using density functional theory(DFT) calculations, we study the intrinsic defects in BaZrS_(3), a representative CP material.We compare Hubbard-U and hybrid functional methods, both of which have been widely used in addressing the band gap problem of semi-local functionals in DFT. We find that tuning the U value to obtain experimental bulk band gap and then using the obtained U value for defect calculations may result in over-localization of defect states. In the hybrid functional calculation, the band gap of BaZrS_(3)can be accurately obtained. We observe the formation of small S-atom clusters in both methods, which tend to self-passivate the defects from forming mid-gap levels. Even though in the hybrid functional calculations several relatively deep defects are observed, all of them exhibit too high formation energy to play a significant role if the materials are prepared under thermal equilibrium.BaZrS_(3)is thus expected to exhibit sufficient defect tolerance promising for photovoltaic and optoelectronic applications.
基金the National Natural Science Foundation of China(11834011 and 11911530142)。
文摘Tin perovskite solar cells(TPSCs)are promising for lead-free perovskite solar cells(PSCs)and have led to extensive research;however,the poor crystallinity and chemical stability of tin perovskites are two issues that prevent stable TPSCs.In this study,we outline a new process that addresses these issues by using tin(II)acetate(Sn(Ac)2)in place of the conventional SnF2 precursor additive.Compared with SnF2,Sn(Ac)2 improves the crystallinity and stability of tin perovskite with fewer defects and better charge extraction.Using this process,we developed a device that has a higher external quantum efficiency for charge extraction compared with the control devices and a power conversion efficiency of 9.93%,which maintained more than 90%of its initial efficiency after 1000 h operation at the maximum power point under standard AM 1.5G solar illumination.
