Aiming at the problem of the low recognition rate of the rolling bearing fault of the walking gearbox of the combine harvester, a gearbox rolling bearing fault diagnosis method based on the dragonfly optimization algorithm kernel extreme learning machine is proposed. The Variational Mode Decomposition(VMD) algorithm optimized by the particle swarm optimization algorithm is used to decompose the experimentally extracted vibration signals of the gearbox in different working states, and the sample entropy value is extracted from the Intrinsic Mode components obtained by the decomposition as the fault characteristic value, and the The time-domain and frequency-domain characteristics of the vibration signal together constitute the fault feature set. The DA-KELM algorithm is used to identify the fault in the feature set of the vibration signal in various states. Through pattern recognition of four states: normal, roller pitting, outer raceway pitting, and inner raceway pitting of the rolling bearing in the traveling gearbox of the combine harvester, The best classification accuracy is 95.625%. At the same time, this method was compared with the common classification algorithm, and the experimental results show that this method has advantages in the accuracy of fault identification.
{"title":"Research on fault diagnosis method of walking gearbox of combine harvester based on DA-KELM","authors":"Zhi Sun, Xinzhong Wang, You Wu","doi":"10.1145/3481113.3481124","DOIUrl":"https://doi.org/10.1145/3481113.3481124","url":null,"abstract":"Aiming at the problem of the low recognition rate of the rolling bearing fault of the walking gearbox of the combine harvester, a gearbox rolling bearing fault diagnosis method based on the dragonfly optimization algorithm kernel extreme learning machine is proposed. The Variational Mode Decomposition(VMD) algorithm optimized by the particle swarm optimization algorithm is used to decompose the experimentally extracted vibration signals of the gearbox in different working states, and the sample entropy value is extracted from the Intrinsic Mode components obtained by the decomposition as the fault characteristic value, and the The time-domain and frequency-domain characteristics of the vibration signal together constitute the fault feature set. The DA-KELM algorithm is used to identify the fault in the feature set of the vibration signal in various states. Through pattern recognition of four states: normal, roller pitting, outer raceway pitting, and inner raceway pitting of the rolling bearing in the traveling gearbox of the combine harvester, The best classification accuracy is 95.625%. At the same time, this method was compared with the common classification algorithm, and the experimental results show that this method has advantages in the accuracy of fault identification.","PeriodicalId":112570,"journal":{"name":"2021 3rd International Symposium on Signal Processing Systems (SSPS)","volume":"279 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-03-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116589012","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
By designing the instrument amplifier, amplifying the signal from the tensiometer and passing it to Mbed, the operator can grasp the relevant information of the weight of the goods. In addition, this article lists several possible options. By weighing their FoMs and prices, the best combination is selected. Block diagram, naive design and accurate design are provided, and the final error analysis is made.
{"title":"Design of Instrument Amplifier Based on Crane","authors":"Haoran Sun","doi":"10.1145/3481113.3481118","DOIUrl":"https://doi.org/10.1145/3481113.3481118","url":null,"abstract":"By designing the instrument amplifier, amplifying the signal from the tensiometer and passing it to Mbed, the operator can grasp the relevant information of the weight of the goods. In addition, this article lists several possible options. By weighing their FoMs and prices, the best combination is selected. Block diagram, naive design and accurate design are provided, and the final error analysis is made.","PeriodicalId":112570,"journal":{"name":"2021 3rd International Symposium on Signal Processing Systems (SSPS)","volume":"5 1-3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-03-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125149202","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This paper introduces a deep learning (DL) framework to address direction of arrival (DOA) estimation problem. Traditional signal processing methods such as multiple signal classification (MUSIC) highly rely on signal model and array geometry. However, DL methods, being data-driven, make analytical process of signal or array less important. In this paper, a neural network architecture combining one-dimensional convolutional neural network (1D CNN) and gated recurrent unit (GRU) is proposed to estimate DOA of multiple signals. The multi-signal DOA estimation is treated as a multi-class multi-label classification issue. First a dataset using the covariance matrix of target signals received by a circular antenna array is generated. The proposed 1D CNN-GRU model then learns the relationship between covariance matrix elements and DOAs through training. Experimental results show that our proposed method has higher accuracy than MUSIC and is able to deal with multi-path DOA estimation. Besides, 1D CNN-GRU is proved to have lower root mean squared error (RMSE) than other DL methods, because features over small local areas and time-sequence are both learnt by 1D CNN layers and GRU layers. In addition, 1D CNN-GRU exhibits effectiveness in experiments using real-world data.
{"title":"Direction of Arrival Estimation Using One-dimensional Convolutional Neural Network and Gated Recurrent Unit","authors":"Mingyue Li, Yougen Xu, Zhiwen Liu","doi":"10.1145/3481113.3481116","DOIUrl":"https://doi.org/10.1145/3481113.3481116","url":null,"abstract":"This paper introduces a deep learning (DL) framework to address direction of arrival (DOA) estimation problem. Traditional signal processing methods such as multiple signal classification (MUSIC) highly rely on signal model and array geometry. However, DL methods, being data-driven, make analytical process of signal or array less important. In this paper, a neural network architecture combining one-dimensional convolutional neural network (1D CNN) and gated recurrent unit (GRU) is proposed to estimate DOA of multiple signals. The multi-signal DOA estimation is treated as a multi-class multi-label classification issue. First a dataset using the covariance matrix of target signals received by a circular antenna array is generated. The proposed 1D CNN-GRU model then learns the relationship between covariance matrix elements and DOAs through training. Experimental results show that our proposed method has higher accuracy than MUSIC and is able to deal with multi-path DOA estimation. Besides, 1D CNN-GRU is proved to have lower root mean squared error (RMSE) than other DL methods, because features over small local areas and time-sequence are both learnt by 1D CNN layers and GRU layers. In addition, 1D CNN-GRU exhibits effectiveness in experiments using real-world data.","PeriodicalId":112570,"journal":{"name":"2021 3rd International Symposium on Signal Processing Systems (SSPS)","volume":"25 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-03-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114819777","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Frequency diverse array (FDA) has angle and range dependent beampattern due to the small frequency increments across array elements, which makes the combination of FDA and multiple input multiple output (MIMO) technique have capability of target localization. However, the introduction of range dimension greatly increases the computational complexity of MUSIC spectrum peak search. To overcome this problem, we propose a closed-form target localization method for FDA-MIMO. First, apply root-MUSIC algorithm on subarray outputs to obtain angle estimation. Then, use rotational invariance between subarray outputs to obtain range estimation with angle estimation obtained before. Simulation results show that compared with the spectrum search method, the proposed method can obtain closed-form solutions with computational amount greatly reduced, and compared with TS-ESPRIT method, the proposed method has higher estimation accuracy.
{"title":"A Closed-Form Target Localization Method for FDA-MIMO Based on root-MUSIC and ESPRIT Algorithm","authors":"Licheng Wang, Yougen Xu, Zhiwen Liu","doi":"10.1145/3481113.3481115","DOIUrl":"https://doi.org/10.1145/3481113.3481115","url":null,"abstract":"Frequency diverse array (FDA) has angle and range dependent beampattern due to the small frequency increments across array elements, which makes the combination of FDA and multiple input multiple output (MIMO) technique have capability of target localization. However, the introduction of range dimension greatly increases the computational complexity of MUSIC spectrum peak search. To overcome this problem, we propose a closed-form target localization method for FDA-MIMO. First, apply root-MUSIC algorithm on subarray outputs to obtain angle estimation. Then, use rotational invariance between subarray outputs to obtain range estimation with angle estimation obtained before. Simulation results show that compared with the spectrum search method, the proposed method can obtain closed-form solutions with computational amount greatly reduced, and compared with TS-ESPRIT method, the proposed method has higher estimation accuracy.","PeriodicalId":112570,"journal":{"name":"2021 3rd International Symposium on Signal Processing Systems (SSPS)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-03-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130727976","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
G. Dimitrov, G. Panayotova, E. Kovatcheva, Pavel Petrov, Magdalena Garvanova, Snejana Petrova, I. Dimitrova, Olexiy Bychkov
In the recent years the attention to Brain-Computer Interface (BCI) devices and their potential for decoding human brain signals have risen considerably. However, a number of issues related to the classification of received signals still remain unresolved. This study focuses on increasing the speed of classification of data obtained from the Brain Computer Interface (BCI), without significantly affecting the accuracy of processing and classification. Our research team focuses on the possibilities to reduce the number of channels as one of the potential factors for increasing the speed of incoming data classification. Experimental data is obtained by using Emotiv Epoc 14+. For data processing we used Python. The data is classified with K-Neighbors algorithm.
{"title":"Decrease the time for classification of the incoming signals from BCI","authors":"G. Dimitrov, G. Panayotova, E. Kovatcheva, Pavel Petrov, Magdalena Garvanova, Snejana Petrova, I. Dimitrova, Olexiy Bychkov","doi":"10.1145/3481113.3481126","DOIUrl":"https://doi.org/10.1145/3481113.3481126","url":null,"abstract":"In the recent years the attention to Brain-Computer Interface (BCI) devices and their potential for decoding human brain signals have risen considerably. However, a number of issues related to the classification of received signals still remain unresolved. This study focuses on increasing the speed of classification of data obtained from the Brain Computer Interface (BCI), without significantly affecting the accuracy of processing and classification. Our research team focuses on the possibilities to reduce the number of channels as one of the potential factors for increasing the speed of incoming data classification. Experimental data is obtained by using Emotiv Epoc 14+. For data processing we used Python. The data is classified with K-Neighbors algorithm.","PeriodicalId":112570,"journal":{"name":"2021 3rd International Symposium on Signal Processing Systems (SSPS)","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-03-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122906434","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Convolutional Neural Network (CNN) is widely used in target tracking for the computer vision, where Intersection of union (IOU) is the most popular evaluation metric in the target detection criteria, but IOU cannot be optimized for tracking algorithms in the case of non-overlapping bounding boxes. GIOU can be optimized for tracking in the case of non-overlapping bounding boxes, but the slow convergence speed of GIOU leads to inaccurate detection, which results in low tracking accuracy. To solve the above problems, a DIOU-based MDNet tracking method is proposed in this paper. In order to solve DIOU loss does not have a penalty term for the aspect ratio of the target box, we propose CIOU-based MDNet and experiments show that the accuracy of this method is improved by 3% compared with MDNet trained with traditional IOU, GIOU or DIOU.
卷积神经网络(Convolutional Neural Network, CNN)广泛应用于计算机视觉的目标跟踪中，其中IOU (Intersection of union)是目标检测标准中最常用的评价指标，但IOU无法优化用于无重叠边界盒情况下的跟踪算法。在不重叠的边界框情况下，可以对GIOU进行跟踪优化，但由于GIOU收敛速度慢，导致检测不准确，导致跟踪精度不高。针对上述问题，本文提出了一种基于diou的MDNet跟踪方法。为了解决DIOU损失对目标框的长宽比没有惩罚项的问题，我们提出了基于ciou的MDNet，实验表明，与传统IOU、GIOU或DIOU训练的MDNet相比，该方法的准确率提高了3%。
{"title":"Multi-domain learning target tracking algorithm based on objective regression optimization","authors":"Xi Yue","doi":"10.1145/3481113.3481122","DOIUrl":"https://doi.org/10.1145/3481113.3481122","url":null,"abstract":"Convolutional Neural Network (CNN) is widely used in target tracking for the computer vision, where Intersection of union (IOU) is the most popular evaluation metric in the target detection criteria, but IOU cannot be optimized for tracking algorithms in the case of non-overlapping bounding boxes. GIOU can be optimized for tracking in the case of non-overlapping bounding boxes, but the slow convergence speed of GIOU leads to inaccurate detection, which results in low tracking accuracy. To solve the above problems, a DIOU-based MDNet tracking method is proposed in this paper. In order to solve DIOU loss does not have a penalty term for the aspect ratio of the target box, we propose CIOU-based MDNet and experiments show that the accuracy of this method is improved by 3% compared with MDNet trained with traditional IOU, GIOU or DIOU.","PeriodicalId":112570,"journal":{"name":"2021 3rd International Symposium on Signal Processing Systems (SSPS)","volume":"24 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-03-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122969155","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In recent years, due to the increasing demand for the understanding and recognition of content in images, image semantic segmentation technology has developed rapidly. Image semantic segmentation technology has also seen more and more reforms and innovations Each classical model has its own innovation and characteristics, which contributes to the development of image semantic segmentation.In this paper, four popular semantic segmentation models are reviewed and their characteristics are introduced.The results show that compared with other models, the SETR model based on Transformer has a higher performance level in semantic segmentation results.
{"title":"Analysis on Approaches and Structures of Image Semantic Segmentation","authors":"Haozheng Ji","doi":"10.1145/3481113.3481123","DOIUrl":"https://doi.org/10.1145/3481113.3481123","url":null,"abstract":"In recent years, due to the increasing demand for the understanding and recognition of content in images, image semantic segmentation technology has developed rapidly. Image semantic segmentation technology has also seen more and more reforms and innovations Each classical model has its own innovation and characteristics, which contributes to the development of image semantic segmentation.In this paper, four popular semantic segmentation models are reviewed and their characteristics are introduced.The results show that compared with other models, the SETR model based on Transformer has a higher performance level in semantic segmentation results.","PeriodicalId":112570,"journal":{"name":"2021 3rd International Symposium on Signal Processing Systems (SSPS)","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-03-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123165343","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The medical image is a set of all organizations, institutions, and resources whose primary goal is to improve health. The extensive growth of medical data increases the utility of machine learning and deep learning in the healthcare domains. Nowadays, the use of in-depth training to process medical images has received particular attention. In recent years, medical instruments have developed rapidly with the help of artificial intelligence and are widely used to process medical images. Artificial intelligence is numerous sources of medical imaging processing such as X-ray, Computed Tomography (CT), and Magnetic Resonance Imaging (MRI). CT and MRI image processing tasks with a high computation time requirement and computation speed. Nowadays, one of the most critical trends in the development of computer technology in neuroscience is the processing of medical images and digital images, which are used to improve image quality, restore damaged images, identify individual elements and diagnose various diseases. In this paper, we briefly review the progress and challenges associated with in-deep learning in the processing of CT and MRI medical images.
{"title":"An Overview of Deep Learning in MRI and CT Medical Image Processing","authors":"Ahliddin Shomirov, Jing Zhang","doi":"10.1145/3481113.3481125","DOIUrl":"https://doi.org/10.1145/3481113.3481125","url":null,"abstract":"The medical image is a set of all organizations, institutions, and resources whose primary goal is to improve health. The extensive growth of medical data increases the utility of machine learning and deep learning in the healthcare domains. Nowadays, the use of in-depth training to process medical images has received particular attention. In recent years, medical instruments have developed rapidly with the help of artificial intelligence and are widely used to process medical images. Artificial intelligence is numerous sources of medical imaging processing such as X-ray, Computed Tomography (CT), and Magnetic Resonance Imaging (MRI). CT and MRI image processing tasks with a high computation time requirement and computation speed. Nowadays, one of the most critical trends in the development of computer technology in neuroscience is the processing of medical images and digital images, which are used to improve image quality, restore damaged images, identify individual elements and diagnose various diseases. In this paper, we briefly review the progress and challenges associated with in-deep learning in the processing of CT and MRI medical images.","PeriodicalId":112570,"journal":{"name":"2021 3rd International Symposium on Signal Processing Systems (SSPS)","volume":"114 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-03-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123356943","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
WFOS (Wide-Field Optical Spectrograph) is a first light instrument mounted on the Nasmyth focus of the TMT (Thirty Meter Telescope). The transient analysis was performed under infrequent earthquakes to study the WFOS shell structure's transient response using Abaqus software. Computer-aided design models of WFOS shell structure were established using Invar (4J32), a nickel-iron alloy, as the selected structural material. Time history analysis was performed on the WFOS structure under seven different earthquake time histories, and the structural responses of displacement, stress, and acceleration were investigated. the peak acceleration is reached when the support structure is dealt against seven infrequent earthquakes in three directions. This research's methodology and outcomes provide a guidance approach to seismic response analysis of the telescope components. A detailed briefing of the computer-aided design model and results are presented.
{"title":"Computational seismic analysis of optical instrument structures: Computational analysis of seismic response of wide-field optical spectrograph structures in large telescopes","authors":"Aman Shrestha, Daxu Zhang, Lingyu Zheng","doi":"10.1145/3481113.3481121","DOIUrl":"https://doi.org/10.1145/3481113.3481121","url":null,"abstract":"WFOS (Wide-Field Optical Spectrograph) is a first light instrument mounted on the Nasmyth focus of the TMT (Thirty Meter Telescope). The transient analysis was performed under infrequent earthquakes to study the WFOS shell structure's transient response using Abaqus software. Computer-aided design models of WFOS shell structure were established using Invar (4J32), a nickel-iron alloy, as the selected structural material. Time history analysis was performed on the WFOS structure under seven different earthquake time histories, and the structural responses of displacement, stress, and acceleration were investigated. the peak acceleration is reached when the support structure is dealt against seven infrequent earthquakes in three directions. This research's methodology and outcomes provide a guidance approach to seismic response analysis of the telescope components. A detailed briefing of the computer-aided design model and results are presented.","PeriodicalId":112570,"journal":{"name":"2021 3rd International Symposium on Signal Processing Systems (SSPS)","volume":"36 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-03-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133316513","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The article shown that for digital signal processing varying in a limited range of amplitudes it is advisable to consider a set of signal levels through its mapping into certain Galois field, i.e., finite commutative body. In this case, the signal coding differs from binary, however, this creates quite definite advantages. In particular, a modified Walsh basis where the elements +1 and -1 are treated as elements of a non-binary Galois field can be used. The main difference of such use of the Walsh basis is that the elements of the Galois field corresponding to the spectral components belong to the same set as the original signal levels do. This provides a significant reduction in the amount of information when transmitting information about the signal, presented in the form of its spectrum. A specific example of using the Galois field for processing a time series of data that simulates a digital signal is presented
{"title":"Spectral representations of digital signals using non-binary Galois fields.","authors":"I. Suleimenov, D. Matrassulova, I. Moldakhan","doi":"10.1145/3481113.3481120","DOIUrl":"https://doi.org/10.1145/3481113.3481120","url":null,"abstract":"The article shown that for digital signal processing varying in a limited range of amplitudes it is advisable to consider a set of signal levels through its mapping into certain Galois field, i.e., finite commutative body. In this case, the signal coding differs from binary, however, this creates quite definite advantages. In particular, a modified Walsh basis where the elements +1 and -1 are treated as elements of a non-binary Galois field can be used. The main difference of such use of the Walsh basis is that the elements of the Galois field corresponding to the spectral components belong to the same set as the original signal levels do. This provides a significant reduction in the amount of information when transmitting information about the signal, presented in the form of its spectrum. A specific example of using the Galois field for processing a time series of data that simulates a digital signal is presented","PeriodicalId":112570,"journal":{"name":"2021 3rd International Symposium on Signal Processing Systems (SSPS)","volume":"12 7","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-03-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"113963684","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：