基于计算机视觉的航拍绝缘子缺陷检测方法被广泛应用于电力巡检。针对绝缘子缺陷易受背景复杂、目标尺度较小等因素的影响而导致漏检、误检的问题,提出了一种旨在提高绝缘子缺陷检测精度的绝缘子缺陷检测模型YOLO-insulator。首先,引入...基于计算机视觉的航拍绝缘子缺陷检测方法被广泛应用于电力巡检。针对绝缘子缺陷易受背景复杂、目标尺度较小等因素的影响而导致漏检、误检的问题,提出了一种旨在提高绝缘子缺陷检测精度的绝缘子缺陷检测模型YOLO-insulator。首先,引入基于通道混洗的重参数化卷积(reparameterized convolution based on channel shuffle-one-shot aggregation, RCS-OSA)替换传统的二维卷积C2f,以增强网络的特征提取能力;其次,在颈部网络使用RCS-OSA模块替换部分的C2f卷积,同时引入挤压激励网络(squeeze and excitation network,SENet),以增强模型对通道间关系的捕捉和整体特征的表达能力;最后,针对多种缺陷区域小导致难以检测的问题,提出小目标检测层方法,该层包含更多的缺陷细节信息,有利于缺陷的检测。在自制绝缘子数据集上进行实验验证的结果表明,相对于基线YOLOv8n,YOLO-insulator模型在查准率、召回率、平均精度均值上都实现了提升,有效提高了模型的综合性能。展开更多
Rod insulators are vital parts of the catenary of high speed railways(HSRs).There are many different catenary insulators,and the background of the insulator image is complicated.It is difficult to recognise insulators...Rod insulators are vital parts of the catenary of high speed railways(HSRs).There are many different catenary insulators,and the background of the insulator image is complicated.It is difficult to recognise insulators and detect defects automatically.In this paper,we propose a catenary intelligent defect detection algorithm based on Mask region-convolutional neural network(R-CNN)and an image processing model.Vertical projection technology is used to achieve single shed positioning and precise cutting of the insulator.Gradient,texture,and gray feature fusion(GTGFF)and a K-means clustering analysis model(KCAM)are proposed to detect broken insulators,dirt,foreign bodies,and flashover.Using this model,insulator recognition and defect detection can achieve a high recall rate and accuracy,and generalized defect detection.The algorithm is tested and verified on a dataset of realistic insulator images,and the accuracy and reliability of the algorithm satisfy current requirements for HSR catenary automatic inspection and intelligent maintenance.展开更多
To address the challenges of high model complexity and low accuracy in insulator component defect detection from drone-captured images,this paper presents adaptive downsampling and frequency-position fusion(ADFP),a li...To address the challenges of high model complexity and low accuracy in insulator component defect detection from drone-captured images,this paper presents adaptive downsampling and frequency-position fusion(ADFP),a lightweight algorithm based on YOLOv11(You Only Look Once version 11).The algorithm presents efficient downsampling module,new feature extraction module and innovative neck structure.By integrating the spatial channel attention module of frequency-aware cascade attention(FCA)and the ADown module,the number of parameters is reduced while accuracy is significantly improved.Additionally,the neck module is redesigned,and the position-aware key feature fusion network(PKFN)module is introduced to further improve feature fusion capabilities.Experiments were conducted on the SAID dataset using the improved model.Compared to the original model,the m AP(0.5)of ADFP shows a 5.3%improvement,while the model parameters are reduced by 12.0%.On other public insulator component defect datasets,these improvements still have better results.Multiple experiments have confirmed the effectiveness of the model and its strong generalization ability.展开更多
Insulator defect detection plays a vital role in maintaining the secure operation of power systems.To address the issues of the difficulty of detecting small objects and missing objects due to the small scale,variable...Insulator defect detection plays a vital role in maintaining the secure operation of power systems.To address the issues of the difficulty of detecting small objects and missing objects due to the small scale,variable scale,and fuzzy edge morphology of insulator defects,we construct an insulator dataset with 1600 samples containing flashovers and breakages.Then a simple and effective surface defect detection method of power line insulators for difficult small objects is proposed.Firstly,a high-resolution featuremap is introduced and a small object prediction layer is added so that the model can detect tiny objects.Secondly,a simplified adaptive spatial feature fusion(SASFF)module is introduced to perform cross-scale spatial fusion to improve adaptability to variable multi-scale features.Finally,we propose an enhanced deformable attention mechanism(EDAM)module.By integrating a gating activation function,the model is further inspired to learn a small number of critical sampling points near reference points.And the module can improve the perception of object morphology.The experimental results indicate that concerning the dataset of flashover and breakage defects,this method improves the performance of YOLOv5,YOLOv7,and YOLOv8.In practical application,it can simply and effectively improve the precision of power line insulator defect detection and reduce missing detection for difficult small objects.展开更多
The insulation aging of cross-linked polyethylene(XLPE)cables is the main reason for the reduction in cable life.There is currently a lack of rapid and effective methods for detecting cable insulation defects in power...The insulation aging of cross-linked polyethylene(XLPE)cables is the main reason for the reduction in cable life.There is currently a lack of rapid and effective methods for detecting cable insulation defects in power-related sectors.To this end,this paper presents a method for identifying insulation defects in XLPE cables based on deep learning algorithms.First,the principle of the harmonic method for detecting cable insulation defects is introduced.Second,the ANSYS software is used to simulate the cable insulation layer containing bubbles,protrusions,and water tree defects,and the effects of each type of defect on the magnetic field strength and eddy loss current of the cable insulation layer are analyzed.Then,a total of 10 characteristic quantities of the total harmonic content and 2nd to 10th harmonic currents are constructed to establish a database of cable insulation defects.Finally,the deep learning algorithm,long short-term memory(LSTM),is used to accurately identify the types of insulation defects in cables.The results indicate that the LSTM algorithm can effectively diagnose and identify insulation defects in cables with an accuracy of 95.83%.展开更多
The requirements for the construction of a new power system inevitably pose significant challenges and changes to the operation and maintenance of the power grid.To ensure the safe and stable operation of ultra-high v...The requirements for the construction of a new power system inevitably pose significant challenges and changes to the operation and maintenance of the power grid.To ensure the safe and stable operation of ultra-high voltage(UHV)transmission equipment,this work reports on the principles and preliminary results of using electroluminescence(EL)-based photon counting(PC)methods for early detection of micro-defects in GIS/GIL insulation spacer.In this study,the impact of voltage,gas pressure,and gas composition on the photon response of insulation is examined.Furthermore,the corresponding relationship between defect status and photon response characteristics is explored,along with the discussion of the EL mechanism and its evolution induced by defects.The research results demonstrate that PC measurement exhibits high sensitivity to variations in millimeter-scale defect size,position,and morphology at lower electric fields before partial discharge(PD)initiation.With this regard,this paper reveals promising prospects for the early detection of micro-defects in UHV transmission equipment using PC measurement-based methods.展开更多
We study the local density of states （LDOS） for electrons scattering off the line edge of an atomic step defect on the surface of a three-dimensional （3D） topological insulator （TI） and the line edge of a finite...We study the local density of states （LDOS） for electrons scattering off the line edge of an atomic step defect on the surface of a three-dimensional （3D） topological insulator （TI） and the line edge of a finite 3D TI, where the front surface and side surface meet with different Fermi velocities, respectively. By using a S-function potential to model the edges, we find that the bound states existed along the step line edge significantly contribute to the LDOS near the edge, but do not modify the exponential behavior away from it. In addition, the power-law decaying behavior for LDOS oscillation away from the step is understood from the spin rotation for surface states scattering off the step defect with magnitude depending on the strength of the potential. Furthermore, the electron refraction and total reflection analogous to optics occurred at the line edge where two surfaces meet with different Fermi velocities, which leads to the LDOS decaying behavior in the greater Fermi velocity side similar to that for a step line edge. However, in the smaller velocity side the LDOS shows a different decaying behavior as x-1/2, and the wavevector of LDOS oscillation is no longer equal to the diameter of the constant energy contour of surface band, but is sensitively dependent on the ratio of the two Fermi velocities. These effects may be verified by STM measurement with high precision.展开更多
Intrinsic magnetic topological insulators have been reported to exhibit novel physical phenomena such as the quantum anomalous Hall effect and axion insulator states,demonstrating potential for applications in spintro...Intrinsic magnetic topological insulators have been reported to exhibit novel physical phenomena such as the quantum anomalous Hall effect and axion insulator states,demonstrating potential for applications in spintronics and topological quantum computing.Here we perform low-temperature scanning tunneling microscopy(STM)investigations of the antiferromagnetic ground state of MnSb_(2)Te_(4),a predicted magnetic topological insulator isostructural with MnBi_(2)Te_(4).We visualize the hexagonal Te-terminated surface of MnSb_(2)Te_(4)and identify two distinct defects originating from different antisite substitutions.Notably,we identify an in-gap state above the Fermi energy where the tunneling spectrum exhibits a negative differential conductance behavior.This electronic state can be modulated by external electric and magnetic fields,suggesting effective pathways for electronic state manipulation.Spin-resolved STM measurements further reveal additional magnetic resonance peaks associated with Mn antisite defects.Our results provide novel insights into the investigation of magnetic topological insulators and demonstrate a promising approach to modulate the localized electronic states.展开更多
针对绝缘子缺陷检测算法具有较大的参数规模和计算量导致难以部署在边缘设备,模型剪枝后难以获得正确连接,且过度稀疏化训练导致模型精度大幅度下降等问题,提出一种基于DepGraph偏移正则化的绝缘子多缺陷检测轻量化算法。通过依赖图(Dep...针对绝缘子缺陷检测算法具有较大的参数规模和计算量导致难以部署在边缘设备,模型剪枝后难以获得正确连接,且过度稀疏化训练导致模型精度大幅度下降等问题,提出一种基于DepGraph偏移正则化的绝缘子多缺陷检测轻量化算法。通过依赖图(DepGraph)对改进后YOLOv7网络建立连接关系模型,再添加偏移正则化稀疏约束对其进行组级的稀疏训练,删除冗余的连接,得到参数规模和计算量更小的轻量型检测算法。将提出的模型压缩算法应用到绝缘子多缺陷检测任务中,实验结果表明,剪枝后模型相较于未剪枝模型的参数规模和计算量分别下降65.25%和65.98%,而平均准确率仅减少1.1个百分点,验证了DepGraph偏移正则化方案在绝缘子多缺陷检测任务中的有效性;在CIFAR-10数据集上进行实验,实验结果表明,在加速比为2.88时,所提算法仍可以保持93.69%的分类精度。使用TensorRT对该算法进行推理加速,并在Jetson Orin Nano平台上部署,经过TensorRT优化后模型的检测速度达到了35.24帧/s,符合在移动设备上部署的需求。展开更多
文摘基于计算机视觉的航拍绝缘子缺陷检测方法被广泛应用于电力巡检。针对绝缘子缺陷易受背景复杂、目标尺度较小等因素的影响而导致漏检、误检的问题,提出了一种旨在提高绝缘子缺陷检测精度的绝缘子缺陷检测模型YOLO-insulator。首先,引入基于通道混洗的重参数化卷积(reparameterized convolution based on channel shuffle-one-shot aggregation, RCS-OSA)替换传统的二维卷积C2f,以增强网络的特征提取能力;其次,在颈部网络使用RCS-OSA模块替换部分的C2f卷积,同时引入挤压激励网络(squeeze and excitation network,SENet),以增强模型对通道间关系的捕捉和整体特征的表达能力;最后,针对多种缺陷区域小导致难以检测的问题,提出小目标检测层方法,该层包含更多的缺陷细节信息,有利于缺陷的检测。在自制绝缘子数据集上进行实验验证的结果表明,相对于基线YOLOv8n,YOLO-insulator模型在查准率、召回率、平均精度均值上都实现了提升,有效提高了模型的综合性能。
基金supported by the National Natural Science Foundation of China(Nos.51677171,51637009,51577166 and 51827810)the National Key R&D Program of China(No.2018YFB0606000)+2 种基金the China Scholarship Council(No.201708330502)the Fund of Shuohuang Railway Development Limited Liability Company(No.SHTL-2020-13)the Fund of State Key Laboratory of Industrial Control Technology(No.ICT2022B29),China。
文摘Rod insulators are vital parts of the catenary of high speed railways(HSRs).There are many different catenary insulators,and the background of the insulator image is complicated.It is difficult to recognise insulators and detect defects automatically.In this paper,we propose a catenary intelligent defect detection algorithm based on Mask region-convolutional neural network(R-CNN)and an image processing model.Vertical projection technology is used to achieve single shed positioning and precise cutting of the insulator.Gradient,texture,and gray feature fusion(GTGFF)and a K-means clustering analysis model(KCAM)are proposed to detect broken insulators,dirt,foreign bodies,and flashover.Using this model,insulator recognition and defect detection can achieve a high recall rate and accuracy,and generalized defect detection.The algorithm is tested and verified on a dataset of realistic insulator images,and the accuracy and reliability of the algorithm satisfy current requirements for HSR catenary automatic inspection and intelligent maintenance.
基金Supported by the Natural Science Foundution of Heilongjiang Province(LH2024E109)。
文摘To address the challenges of high model complexity and low accuracy in insulator component defect detection from drone-captured images,this paper presents adaptive downsampling and frequency-position fusion(ADFP),a lightweight algorithm based on YOLOv11(You Only Look Once version 11).The algorithm presents efficient downsampling module,new feature extraction module and innovative neck structure.By integrating the spatial channel attention module of frequency-aware cascade attention(FCA)and the ADown module,the number of parameters is reduced while accuracy is significantly improved.Additionally,the neck module is redesigned,and the position-aware key feature fusion network(PKFN)module is introduced to further improve feature fusion capabilities.Experiments were conducted on the SAID dataset using the improved model.Compared to the original model,the m AP(0.5)of ADFP shows a 5.3%improvement,while the model parameters are reduced by 12.0%.On other public insulator component defect datasets,these improvements still have better results.Multiple experiments have confirmed the effectiveness of the model and its strong generalization ability.
基金State Grid Jiangsu Electric Power Co.,Ltd.of the Science and Technology Project(Grant No.J2022004).
文摘Insulator defect detection plays a vital role in maintaining the secure operation of power systems.To address the issues of the difficulty of detecting small objects and missing objects due to the small scale,variable scale,and fuzzy edge morphology of insulator defects,we construct an insulator dataset with 1600 samples containing flashovers and breakages.Then a simple and effective surface defect detection method of power line insulators for difficult small objects is proposed.Firstly,a high-resolution featuremap is introduced and a small object prediction layer is added so that the model can detect tiny objects.Secondly,a simplified adaptive spatial feature fusion(SASFF)module is introduced to perform cross-scale spatial fusion to improve adaptability to variable multi-scale features.Finally,we propose an enhanced deformable attention mechanism(EDAM)module.By integrating a gating activation function,the model is further inspired to learn a small number of critical sampling points near reference points.And the module can improve the perception of object morphology.The experimental results indicate that concerning the dataset of flashover and breakage defects,this method improves the performance of YOLOv5,YOLOv7,and YOLOv8.In practical application,it can simply and effectively improve the precision of power line insulator defect detection and reduce missing detection for difficult small objects.
基金supported by the technology project of the State Grid Shanxi Electric Power Company.The name of the project is“Research and Application of Cable electrification diagnosis Technology based on Harmonic method”(5205C02000GL).
文摘The insulation aging of cross-linked polyethylene(XLPE)cables is the main reason for the reduction in cable life.There is currently a lack of rapid and effective methods for detecting cable insulation defects in power-related sectors.To this end,this paper presents a method for identifying insulation defects in XLPE cables based on deep learning algorithms.First,the principle of the harmonic method for detecting cable insulation defects is introduced.Second,the ANSYS software is used to simulate the cable insulation layer containing bubbles,protrusions,and water tree defects,and the effects of each type of defect on the magnetic field strength and eddy loss current of the cable insulation layer are analyzed.Then,a total of 10 characteristic quantities of the total harmonic content and 2nd to 10th harmonic currents are constructed to establish a database of cable insulation defects.Finally,the deep learning algorithm,long short-term memory(LSTM),is used to accurately identify the types of insulation defects in cables.The results indicate that the LSTM algorithm can effectively diagnose and identify insulation defects in cables with an accuracy of 95.83%.
基金supported by the National Natural Science Foundation of China under Grant No.52125703.
文摘The requirements for the construction of a new power system inevitably pose significant challenges and changes to the operation and maintenance of the power grid.To ensure the safe and stable operation of ultra-high voltage(UHV)transmission equipment,this work reports on the principles and preliminary results of using electroluminescence(EL)-based photon counting(PC)methods for early detection of micro-defects in GIS/GIL insulation spacer.In this study,the impact of voltage,gas pressure,and gas composition on the photon response of insulation is examined.Furthermore,the corresponding relationship between defect status and photon response characteristics is explored,along with the discussion of the EL mechanism and its evolution induced by defects.The research results demonstrate that PC measurement exhibits high sensitivity to variations in millimeter-scale defect size,position,and morphology at lower electric fields before partial discharge(PD)initiation.With this regard,this paper reveals promising prospects for the early detection of micro-defects in UHV transmission equipment using PC measurement-based methods.
基金Project supported by the National Natural Science Foundation of China(Grant No.11274108)the Specialized Research Fund for the Doctoral Program of Higher Education of China(Grant No.20114306110008)the Hunan Provincial Innovation Foundation for Postgraduates(Grant No.CX2012B204)
文摘We study the local density of states （LDOS） for electrons scattering off the line edge of an atomic step defect on the surface of a three-dimensional （3D） topological insulator （TI） and the line edge of a finite 3D TI, where the front surface and side surface meet with different Fermi velocities, respectively. By using a S-function potential to model the edges, we find that the bound states existed along the step line edge significantly contribute to the LDOS near the edge, but do not modify the exponential behavior away from it. In addition, the power-law decaying behavior for LDOS oscillation away from the step is understood from the spin rotation for surface states scattering off the step defect with magnitude depending on the strength of the potential. Furthermore, the electron refraction and total reflection analogous to optics occurred at the line edge where two surfaces meet with different Fermi velocities, which leads to the LDOS decaying behavior in the greater Fermi velocity side similar to that for a step line edge. However, in the smaller velocity side the LDOS shows a different decaying behavior as x-1/2, and the wavevector of LDOS oscillation is no longer equal to the diameter of the constant energy contour of surface band, but is sensitively dependent on the ratio of the two Fermi velocities. These effects may be verified by STM measurement with high precision.
基金Project supported by the National Key R&D Program of China(Grant Nos.2022YFA1403800 and 2023YFA1406500)the National Natural Science Foundation of China(Grant No.12274459)。
文摘Intrinsic magnetic topological insulators have been reported to exhibit novel physical phenomena such as the quantum anomalous Hall effect and axion insulator states,demonstrating potential for applications in spintronics and topological quantum computing.Here we perform low-temperature scanning tunneling microscopy(STM)investigations of the antiferromagnetic ground state of MnSb_(2)Te_(4),a predicted magnetic topological insulator isostructural with MnBi_(2)Te_(4).We visualize the hexagonal Te-terminated surface of MnSb_(2)Te_(4)and identify two distinct defects originating from different antisite substitutions.Notably,we identify an in-gap state above the Fermi energy where the tunneling spectrum exhibits a negative differential conductance behavior.This electronic state can be modulated by external electric and magnetic fields,suggesting effective pathways for electronic state manipulation.Spin-resolved STM measurements further reveal additional magnetic resonance peaks associated with Mn antisite defects.Our results provide novel insights into the investigation of magnetic topological insulators and demonstrate a promising approach to modulate the localized electronic states.
文摘针对绝缘子缺陷检测算法具有较大的参数规模和计算量导致难以部署在边缘设备,模型剪枝后难以获得正确连接,且过度稀疏化训练导致模型精度大幅度下降等问题,提出一种基于DepGraph偏移正则化的绝缘子多缺陷检测轻量化算法。通过依赖图(DepGraph)对改进后YOLOv7网络建立连接关系模型,再添加偏移正则化稀疏约束对其进行组级的稀疏训练,删除冗余的连接,得到参数规模和计算量更小的轻量型检测算法。将提出的模型压缩算法应用到绝缘子多缺陷检测任务中,实验结果表明,剪枝后模型相较于未剪枝模型的参数规模和计算量分别下降65.25%和65.98%,而平均准确率仅减少1.1个百分点,验证了DepGraph偏移正则化方案在绝缘子多缺陷检测任务中的有效性;在CIFAR-10数据集上进行实验,实验结果表明,在加速比为2.88时,所提算法仍可以保持93.69%的分类精度。使用TensorRT对该算法进行推理加速,并在Jetson Orin Nano平台上部署,经过TensorRT优化后模型的检测速度达到了35.24帧/s,符合在移动设备上部署的需求。
