Pub Date : 2022-02-01DOI: 10.1109/ICTech55460.2022.00050
Xiaolong Liu, Xiaojun Xia, Jiaqiang Song
When a rolling bearing fails, the vibration signal of the bearing is unstable and the signal presents non-linear characteristics. As a result, the existing rolling bearing fault diagnosis system has a weak ability to extract the original signal, and the poor ability to identify the rolling bearing signal leads to the final diagnosis effect and expected performance. There is a big gap, in order to enhance the intelligence of the fault diagnosis system, improve the accuracy and generalization ability of the system, and adapt to the needs of factory big data fault diagnosis. This paper proposes a fault diagnosis method of rolling bearing based on improved convolution neural network. First, this method improves the existing activation function and pooling method. After the convolutional layer and pooling, a layer of convolutional layer is added, and the stochastic gradient descent algorithm is used to accelerate the training speed. At the same time, an improved uniformity is proposed. The variance is used as the loss function of the network. The method proposed in this paper is experimentally verified under the bearing data set of Case Western Reserve University, the classic rolling bearing data set, and the conclusion is drawn through the experiment: the experiment under the bearing data set of Case Western Reserve University of the classic rolling bearing data set has achieved better results than the traditional The model has better experimental results, good anti-dryness and better generalization ability. This diagnosis method provides a new idea for fault diagnosis methods, and has a good technical application prospect in industrial production.
{"title":"Research on Fault Diagnosis Method of Rolling Bearing Based on Improved Convolutional Neural Network","authors":"Xiaolong Liu, Xiaojun Xia, Jiaqiang Song","doi":"10.1109/ICTech55460.2022.00050","DOIUrl":"https://doi.org/10.1109/ICTech55460.2022.00050","url":null,"abstract":"When a rolling bearing fails, the vibration signal of the bearing is unstable and the signal presents non-linear characteristics. As a result, the existing rolling bearing fault diagnosis system has a weak ability to extract the original signal, and the poor ability to identify the rolling bearing signal leads to the final diagnosis effect and expected performance. There is a big gap, in order to enhance the intelligence of the fault diagnosis system, improve the accuracy and generalization ability of the system, and adapt to the needs of factory big data fault diagnosis. This paper proposes a fault diagnosis method of rolling bearing based on improved convolution neural network. First, this method improves the existing activation function and pooling method. After the convolutional layer and pooling, a layer of convolutional layer is added, and the stochastic gradient descent algorithm is used to accelerate the training speed. At the same time, an improved uniformity is proposed. The variance is used as the loss function of the network. The method proposed in this paper is experimentally verified under the bearing data set of Case Western Reserve University, the classic rolling bearing data set, and the conclusion is drawn through the experiment: the experiment under the bearing data set of Case Western Reserve University of the classic rolling bearing data set has achieved better results than the traditional The model has better experimental results, good anti-dryness and better generalization ability. This diagnosis method provides a new idea for fault diagnosis methods, and has a good technical application prospect in industrial production.","PeriodicalId":290836,"journal":{"name":"2022 11th International Conference of Information and Communication Technology (ICTech))","volume":"217 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122884000","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}
Pub Date : 2022-02-01DOI: 10.1109/ICTech55460.2022.00017
Yiying Zhang, Caixia Ma, Yeshen He, Kun Liang, Yannian Wu, Zhu Liu
In natural language processing, intelligent question answering based on knowledge graph has received great attention. In the previous knowledge base question answering, the traditional word vector is difficult to express the text semantic information, and the cyclic neural network is easy to cause gradient disappearance and gradient explosion. At the same time, it is lack of comprehensive consideration of text context information. This paper proposes an intelligent Q & A method based on knowledge graph, which uses BiLSTM-CRF model to realize entity recognition. The intelligent Q & A model is constructed based on Seq2Seq, and the above methods are verified by taking the intelligent Q & A as an example, which effectively improves the accuracy of intelligent Q & A.
{"title":"A Question Answering Method of Knowledge Graph Based on BiLSTM-CRF and Seq2Seq","authors":"Yiying Zhang, Caixia Ma, Yeshen He, Kun Liang, Yannian Wu, Zhu Liu","doi":"10.1109/ICTech55460.2022.00017","DOIUrl":"https://doi.org/10.1109/ICTech55460.2022.00017","url":null,"abstract":"In natural language processing, intelligent question answering based on knowledge graph has received great attention. In the previous knowledge base question answering, the traditional word vector is difficult to express the text semantic information, and the cyclic neural network is easy to cause gradient disappearance and gradient explosion. At the same time, it is lack of comprehensive consideration of text context information. This paper proposes an intelligent Q & A method based on knowledge graph, which uses BiLSTM-CRF model to realize entity recognition. The intelligent Q & A model is constructed based on Seq2Seq, and the above methods are verified by taking the intelligent Q & A as an example, which effectively improves the accuracy of intelligent Q & A.","PeriodicalId":290836,"journal":{"name":"2022 11th International Conference of Information and Communication Technology (ICTech))","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123919766","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}
Pub Date : 2022-02-01DOI: 10.1109/ICTech55460.2022.00039
Jie Jiao, Bihui Yu, Huajun Sun
In recent years, in the context of the vigorous development of knowledge graphs, the data scale of the related Semantic Web has also shown an explosive trend. This makes the connection between semantic data more abundant. The RDFS rules and OWL rules currently used in the mainstream of data reasoning in the Semantic Web. When reasoning on large-scale data, the general method is to use forward flow parallel reasoning. In this process, due to the limitation of the reasoning rule set, the more common OWL Horst rule set often makes the content of reasoning insufficient. In this paper, the standard query language SPARQL of the Semantic Web is used to realize the design and implementation of a conversion method corresponding to OWL axioms and OWL Horst rule sets, so as to expand the forward flow reasoning ability based on OWL Horst rules. Through data LUBM and existing Experiments on the reasoning algorithm of this method have verified the feasibility of this method.
{"title":"Forward Reasoning of Owl Rule Set Based on SPARQL Query","authors":"Jie Jiao, Bihui Yu, Huajun Sun","doi":"10.1109/ICTech55460.2022.00039","DOIUrl":"https://doi.org/10.1109/ICTech55460.2022.00039","url":null,"abstract":"In recent years, in the context of the vigorous development of knowledge graphs, the data scale of the related Semantic Web has also shown an explosive trend. This makes the connection between semantic data more abundant. The RDFS rules and OWL rules currently used in the mainstream of data reasoning in the Semantic Web. When reasoning on large-scale data, the general method is to use forward flow parallel reasoning. In this process, due to the limitation of the reasoning rule set, the more common OWL Horst rule set often makes the content of reasoning insufficient. In this paper, the standard query language SPARQL of the Semantic Web is used to realize the design and implementation of a conversion method corresponding to OWL axioms and OWL Horst rule sets, so as to expand the forward flow reasoning ability based on OWL Horst rules. Through data LUBM and existing Experiments on the reasoning algorithm of this method have verified the feasibility of this method.","PeriodicalId":290836,"journal":{"name":"2022 11th International Conference of Information and Communication Technology (ICTech))","volume":"51 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121230356","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}
Pub Date : 2022-02-01DOI: 10.1109/ICTech55460.2022.00035
Li Lv, Haichao Du
With the rapid development of Internet technology, caching technology has become a key technology of various large and medium-sized websites. The quality of cache design is directly related to the speed of website access and the number of servers purchased., and even affects the user experience. With the increase of user access and concurrency, the design of traditional small Internet architecture can not meet the needs of business., and cache technology is particularly important. How to design a cache architecture that can resist high concurrency and large amount of data is a problem worthy of in-depth research. This paper first introduces the commonly used cache technology, then puts forward a scheme of multi-layer cache architecture, analyzes the problems existing in high concurrency., and puts forward the corresponding countermeasures. The cache architecture is further optimized. Finally., the cache architecture is applied to the actual scene of the palm life shopping mall module of the bank app, design the whole architecture and describe the related technology., and then the group experiment is carried out., the pressure test is verified. According to the experimental results., the cache architecture has the characteristics of high availability, anti high concurrency and ensuring data accuracy and stability.
{"title":"Research and Application on Distributed Multi-Level Cache Architecture","authors":"Li Lv, Haichao Du","doi":"10.1109/ICTech55460.2022.00035","DOIUrl":"https://doi.org/10.1109/ICTech55460.2022.00035","url":null,"abstract":"With the rapid development of Internet technology, caching technology has become a key technology of various large and medium-sized websites. The quality of cache design is directly related to the speed of website access and the number of servers purchased., and even affects the user experience. With the increase of user access and concurrency, the design of traditional small Internet architecture can not meet the needs of business., and cache technology is particularly important. How to design a cache architecture that can resist high concurrency and large amount of data is a problem worthy of in-depth research. This paper first introduces the commonly used cache technology, then puts forward a scheme of multi-layer cache architecture, analyzes the problems existing in high concurrency., and puts forward the corresponding countermeasures. The cache architecture is further optimized. Finally., the cache architecture is applied to the actual scene of the palm life shopping mall module of the bank app, design the whole architecture and describe the related technology., and then the group experiment is carried out., the pressure test is verified. According to the experimental results., the cache architecture has the characteristics of high availability, anti high concurrency and ensuring data accuracy and stability.","PeriodicalId":290836,"journal":{"name":"2022 11th International Conference of Information and Communication Technology (ICTech))","volume":"52 1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115979464","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}
Pub Date : 2022-02-01DOI: 10.1109/ICTech55460.2022.00103
Bihui Yu, Chen Deng, Liping Bu
With the development of the Internet, the policy text classification model of deep learning is used to improve the effect of policy text classification, and to play and use the huge value contained in the policy text. In order to more accurately determine the policy field described by the text, a BERT-based policy text classification algorithm is proposed. First, the algorithm uses the BERT (Bidirectional Encoder Representations from Transformers) pre-trained language model to vectorize the sentence-level feature of the policy field text, and then the obtained feature vector is input into the classifier for classification, and finally the homology policy field is used. The text data set is verified. The experimental results show that the classification of the trained model on the test set recorded the highest F1 value of 93.25%. It is nearly 6% higher than the classification task of the BERT model for the MRPC task. Therefore, the proposed policy domain text classification algorithm can more accurately and efficiently judge the domain of the policy text, which is helpful for further analysis of the text data in the policy domain and extract more valuable information.
随着互联网的发展，利用深度学习的政策文本分类模型来提高政策文本分类的效果，发挥和利用政策文本所蕴含的巨大价值。为了更准确地确定文本所描述的策略域，提出了一种基于bert的策略文本分类算法。该算法首先使用BERT (Bidirectional Encoder Representations from Transformers)预训练的语言模型对策略域文本的句子级特征进行矢量化，然后将得到的特征向量输入到分类器中进行分类，最后使用同源策略域。验证文本数据集。实验结果表明，训练后的模型在测试集上的分类F1值最高，达到93.25%。对于MRPC任务，它比BERT模型的分类任务高出近6%。因此，本文提出的策略领域文本分类算法能够更加准确、高效地判断策略文本所属的领域，有助于进一步分析策略领域的文本数据，提取更多有价值的信息。
{"title":"Policy Text Classification Algorithm Based on Bert","authors":"Bihui Yu, Chen Deng, Liping Bu","doi":"10.1109/ICTech55460.2022.00103","DOIUrl":"https://doi.org/10.1109/ICTech55460.2022.00103","url":null,"abstract":"With the development of the Internet, the policy text classification model of deep learning is used to improve the effect of policy text classification, and to play and use the huge value contained in the policy text. In order to more accurately determine the policy field described by the text, a BERT-based policy text classification algorithm is proposed. First, the algorithm uses the BERT (Bidirectional Encoder Representations from Transformers) pre-trained language model to vectorize the sentence-level feature of the policy field text, and then the obtained feature vector is input into the classifier for classification, and finally the homology policy field is used. The text data set is verified. The experimental results show that the classification of the trained model on the test set recorded the highest F1 value of 93.25%. It is nearly 6% higher than the classification task of the BERT model for the MRPC task. Therefore, the proposed policy domain text classification algorithm can more accurately and efficiently judge the domain of the policy text, which is helpful for further analysis of the text data in the policy domain and extract more valuable information.","PeriodicalId":290836,"journal":{"name":"2022 11th International Conference of Information and Communication Technology (ICTech))","volume":"243 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115833598","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}
Pub Date : 2022-02-01DOI: 10.1109/ICTech55460.2022.00041
Zhenyu Yin, Feiqing Zhang, Jiangbo Wang, Fulong Xu, Chao Fan
In recent years, deep learning has been gradually applied to the industry with great success. As the demand for the lightweight intelligent devices increases, the deployment of deep learning models on embedded platforms to meet users' needs for real-time performance has become a trend in the development of intelligence. However, due to the pursuit of higher accuracy, existing deep learning frameworks are becoming richer in functionality and more complex in computation. A large amount of memory requirements and computational power demands make it challenging to deploy neural network computing frameworks on embedded platforms with limited resources and computational power. The WPOC algorithm is proposed and integrated into the Darknet framework to address real-time image processing based on the Winograd algorithm. Tested on the ZYNQ-7010 platform was passed. The results show that the WPOC algorithm proposed in this paper can effectively speed up image recognition by about six times under the VGG-16 model while ensuring the same accuracy rate.
{"title":"Design and Implementation of a Fast Convolution Algorithm for Embedded Platform","authors":"Zhenyu Yin, Feiqing Zhang, Jiangbo Wang, Fulong Xu, Chao Fan","doi":"10.1109/ICTech55460.2022.00041","DOIUrl":"https://doi.org/10.1109/ICTech55460.2022.00041","url":null,"abstract":"In recent years, deep learning has been gradually applied to the industry with great success. As the demand for the lightweight intelligent devices increases, the deployment of deep learning models on embedded platforms to meet users' needs for real-time performance has become a trend in the development of intelligence. However, due to the pursuit of higher accuracy, existing deep learning frameworks are becoming richer in functionality and more complex in computation. A large amount of memory requirements and computational power demands make it challenging to deploy neural network computing frameworks on embedded platforms with limited resources and computational power. The WPOC algorithm is proposed and integrated into the Darknet framework to address real-time image processing based on the Winograd algorithm. Tested on the ZYNQ-7010 platform was passed. The results show that the WPOC algorithm proposed in this paper can effectively speed up image recognition by about six times under the VGG-16 model while ensuring the same accuracy rate.","PeriodicalId":290836,"journal":{"name":"2022 11th International Conference of Information and Communication Technology (ICTech))","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125368887","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}
Pub Date : 2022-02-01DOI: 10.1109/ICTech55460.2022.00110
Yibing Chen, Yuchen He, Hui Chen, Huaibin Zheng
Intensity interferometry (II), as a lensless imaging technique, has the advantages of high resolution. However, the phase retrieval algorithm has the defect of unstable convergence, which leads to partial optimization of II. The introduction of ptychographical into II is considered to overcome this shortcoming, namely ptychographical intensity interferometry imaging (PIII). In this paper, we studied the factors that affect PIII through simulation demonstrations, and the results show that the increase of speckles, overlap and iterations can effectively improve the quality of PIII. This result has guiding significance for the subsequent application of PIII in the field of optics.
{"title":"Analysis of Factors Affecting Ptychographical Intensity Interferometry Imaging","authors":"Yibing Chen, Yuchen He, Hui Chen, Huaibin Zheng","doi":"10.1109/ICTech55460.2022.00110","DOIUrl":"https://doi.org/10.1109/ICTech55460.2022.00110","url":null,"abstract":"Intensity interferometry (II), as a lensless imaging technique, has the advantages of high resolution. However, the phase retrieval algorithm has the defect of unstable convergence, which leads to partial optimization of II. The introduction of ptychographical into II is considered to overcome this shortcoming, namely ptychographical intensity interferometry imaging (PIII). In this paper, we studied the factors that affect PIII through simulation demonstrations, and the results show that the increase of speckles, overlap and iterations can effectively improve the quality of PIII. This result has guiding significance for the subsequent application of PIII in the field of optics.","PeriodicalId":290836,"journal":{"name":"2022 11th International Conference of Information and Communication Technology (ICTech))","volume":"263 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124295549","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}
Pub Date : 2022-02-01DOI: 10.1109/ictech55460.2022.00003
{"title":"Copyright and Reprint Permissions","authors":"","doi":"10.1109/ictech55460.2022.00003","DOIUrl":"https://doi.org/10.1109/ictech55460.2022.00003","url":null,"abstract":"","PeriodicalId":290836,"journal":{"name":"2022 11th International Conference of Information and Communication Technology (ICTech))","volume":"95 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123766714","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}
Pub Date : 2022-02-01DOI: 10.1109/ictech55460.2022.00006
{"title":"Editorial: ICTech 2022","authors":"","doi":"10.1109/ictech55460.2022.00006","DOIUrl":"https://doi.org/10.1109/ictech55460.2022.00006","url":null,"abstract":"","PeriodicalId":290836,"journal":{"name":"2022 11th International Conference of Information and Communication Technology (ICTech))","volume":"100 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125535837","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}
Pub Date : 2022-02-01DOI: 10.1109/ICTech55460.2022.00059
Hongliang Wang, Ruiqi Zhu, Bihui Yu
In recent years, image detection services based on cloud computing deep learning have emerged at the historic moment, but due to the influence of network instability, bandwidth restrictions and many other factors, there may be a large response delay, which will seriously affect the user experience. How to allocate large-scale services to limited nodes, increase user satisfaction, and achieve the load balance, this has become a difficult problem to be solved at present. In this paper, the simulation environment is configured based on cloudsim, and the simulation experiments of standard particle swarm optimization algorithm and improved algorithm are carried out to simulate the scheduling strategy of sensitive image detection service suitable for the deep learning framework of this subject.
{"title":"Research on Sensitive Image Detection Service Based on Deep Learning Framework","authors":"Hongliang Wang, Ruiqi Zhu, Bihui Yu","doi":"10.1109/ICTech55460.2022.00059","DOIUrl":"https://doi.org/10.1109/ICTech55460.2022.00059","url":null,"abstract":"In recent years, image detection services based on cloud computing deep learning have emerged at the historic moment, but due to the influence of network instability, bandwidth restrictions and many other factors, there may be a large response delay, which will seriously affect the user experience. How to allocate large-scale services to limited nodes, increase user satisfaction, and achieve the load balance, this has become a difficult problem to be solved at present. In this paper, the simulation environment is configured based on cloudsim, and the simulation experiments of standard particle swarm optimization algorithm and improved algorithm are carried out to simulate the scheduling strategy of sensitive image detection service suitable for the deep learning framework of this subject.","PeriodicalId":290836,"journal":{"name":"2022 11th International Conference of Information and Communication Technology (ICTech))","volume":"83 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132765764","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：