Pub Date : 2017-11-01DOI: 10.1109/ICSAI.2017.8248338
F. Yang, Lianyi Zhang
Human activity information can be used in a lot of applications such as fitness monitoring. In this paper an activity classification wearable device system is designed which can provide the activity information of the users via a three-axis kinematic sensor. The features in time domain and frequency domain of acceleration data are extracted, and Decision Tree algorithm is applied. The training module is performed once off line to generate the classification model, while the classification can be executed real time on the STM32L low-power microcontroller. The wearable device was worn in watch style in experiment and it offered activity information in acceptable accuracy.
{"title":"Real-time human activity classification by accelerometer embedded wearable devices","authors":"F. Yang, Lianyi Zhang","doi":"10.1109/ICSAI.2017.8248338","DOIUrl":"https://doi.org/10.1109/ICSAI.2017.8248338","url":null,"abstract":"Human activity information can be used in a lot of applications such as fitness monitoring. In this paper an activity classification wearable device system is designed which can provide the activity information of the users via a three-axis kinematic sensor. The features in time domain and frequency domain of acceleration data are extracted, and Decision Tree algorithm is applied. The training module is performed once off line to generate the classification model, while the classification can be executed real time on the STM32L low-power microcontroller. The wearable device was worn in watch style in experiment and it offered activity information in acceptable accuracy.","PeriodicalId":285726,"journal":{"name":"2017 4th International Conference on Systems and Informatics (ICSAI)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116933206","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 : 2017-11-01DOI: 10.1109/ICSAI.2017.8248528
Yunsong Zhong, Qinpei Zhao, Weixiong Rao
With the rapid development of economy, people need to raise the accuracy of predicting prices considering late events. The main tackle in raising the accuracy is to fully use the information in daily reports. Unfortunately, most of the current solutions separate the time series statistics, the events reported in text data and prediction in data mining way from each other. As a result, although they predict the prices accurately in most of the days, they can not grasp the sudden change points of those time series. In this paper, we propose an ensemble framework to take advantage of the news text in predicting change points in stock market indexes as well as traditional prediction works, so that we can improve our prediction sufficiently. Our extensive experimental results shows that we reduce the loss of error predictions and enhance the good prediction results.
{"title":"Predicting stock market indexes with world news","authors":"Yunsong Zhong, Qinpei Zhao, Weixiong Rao","doi":"10.1109/ICSAI.2017.8248528","DOIUrl":"https://doi.org/10.1109/ICSAI.2017.8248528","url":null,"abstract":"With the rapid development of economy, people need to raise the accuracy of predicting prices considering late events. The main tackle in raising the accuracy is to fully use the information in daily reports. Unfortunately, most of the current solutions separate the time series statistics, the events reported in text data and prediction in data mining way from each other. As a result, although they predict the prices accurately in most of the days, they can not grasp the sudden change points of those time series. In this paper, we propose an ensemble framework to take advantage of the news text in predicting change points in stock market indexes as well as traditional prediction works, so that we can improve our prediction sufficiently. Our extensive experimental results shows that we reduce the loss of error predictions and enhance the good prediction results.","PeriodicalId":285726,"journal":{"name":"2017 4th International Conference on Systems and Informatics (ICSAI)","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117146603","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 : 2017-11-01DOI: 10.1109/ICSAI.2017.8248323
Wenjun Xu, Jian Zhang, Yifeng Chen, Wei Xue, Yu Cao
Icing is a major security risk for transmission line of power systems in the cold winter. Traditional deicing methods such as mechanical deicing and thermos-fuse methods are characterized by high energy cost, time consuming and inevitable obvious damage to the external surface of transmission line. A novel method of laser induced bubble explosion reaction (LIBER) to remove ice accretion on transmission line is promoted in this work. The effects of laser power, scanning speed, thickness of foaming layer and ice layer on laser deicing were studied and the results showed that the removal of ice accretion could be achieved with scanning speed of 10 ∼ 20mm/s and laser power ranging from 1000 to 2000W. Ice accretions with the thickness from 3 to 7mm were successfully removed with no damage or negligible damage to the substrate surface, which validated the feasibility of this new LIBER-based deicing method that would have significant applications to the actual deicing scenarios of power grid.
{"title":"Deicing of aluminum alloy rod by laser induced bubble explosion reaction","authors":"Wenjun Xu, Jian Zhang, Yifeng Chen, Wei Xue, Yu Cao","doi":"10.1109/ICSAI.2017.8248323","DOIUrl":"https://doi.org/10.1109/ICSAI.2017.8248323","url":null,"abstract":"Icing is a major security risk for transmission line of power systems in the cold winter. Traditional deicing methods such as mechanical deicing and thermos-fuse methods are characterized by high energy cost, time consuming and inevitable obvious damage to the external surface of transmission line. A novel method of laser induced bubble explosion reaction (LIBER) to remove ice accretion on transmission line is promoted in this work. The effects of laser power, scanning speed, thickness of foaming layer and ice layer on laser deicing were studied and the results showed that the removal of ice accretion could be achieved with scanning speed of 10 ∼ 20mm/s and laser power ranging from 1000 to 2000W. Ice accretions with the thickness from 3 to 7mm were successfully removed with no damage or negligible damage to the substrate surface, which validated the feasibility of this new LIBER-based deicing method that would have significant applications to the actual deicing scenarios of power grid.","PeriodicalId":285726,"journal":{"name":"2017 4th International Conference on Systems and Informatics (ICSAI)","volume":"296 ","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120972363","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 research of neural networks compression becomes a hot spot in the AI area. In this paper, we propose a novel method to optimize the hyperparameters of a compression Neural Networks called HashedNets with Radial Basis Function (RBF) interpolation model and Dynamic Coordinate Search (DYCORS) method, the proposed model is called HD-HORD which can help the HashedNets search for the best hyperparameters automatically and efficiently. Experimental results show that the efficiency of HD-HORD can be improved 72% faster than other methods.
{"title":"A novel HashedNets model based on the efficient hyperparameter optimization","authors":"Qin Fang, Jianxia Chen, Zhongbao Ma, Chao Li, Jie Zhang, Yixin Chen, Qiang Lv","doi":"10.1109/ICSAI.2017.8248458","DOIUrl":"https://doi.org/10.1109/ICSAI.2017.8248458","url":null,"abstract":"The research of neural networks compression becomes a hot spot in the AI area. In this paper, we propose a novel method to optimize the hyperparameters of a compression Neural Networks called HashedNets with Radial Basis Function (RBF) interpolation model and Dynamic Coordinate Search (DYCORS) method, the proposed model is called HD-HORD which can help the HashedNets search for the best hyperparameters automatically and efficiently. Experimental results show that the efficiency of HD-HORD can be improved 72% faster than other methods.","PeriodicalId":285726,"journal":{"name":"2017 4th International Conference on Systems and Informatics (ICSAI)","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124884129","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 : 2017-11-01DOI: 10.1109/ICSAI.2017.8248429
M. R. Silva, P. Machado, L. E. Souza, Carlos Waldecir de Souza
Some performance parameters in IEC 61850 related to the time in processing communication are defined, and these parameters are really important and critical in the automation of power systems. Related to the security analysis of the communication networks, a really powerful tool is the Stochastic Colored Petri Net (SCPN), because it models asynchronous and concurrent processes, and it is useful to analyze delays in timed systems. So, based in this context and knowing that time is really critical in transmitting messages in the IEC 61850 context; it is modeled in an IEC 61850 scenario a cyber-attack using SCPN that aims to delay the transmission of the messages, what is really critical for the operation of the system. The results corresponded to the modeled cyber-attack, showing the efficacy of the proposed modeling method.
{"title":"Modeling of a cyber-attack in an IEC 61850 scenario using stochastic colored Petri Nets","authors":"M. R. Silva, P. Machado, L. E. Souza, Carlos Waldecir de Souza","doi":"10.1109/ICSAI.2017.8248429","DOIUrl":"https://doi.org/10.1109/ICSAI.2017.8248429","url":null,"abstract":"Some performance parameters in IEC 61850 related to the time in processing communication are defined, and these parameters are really important and critical in the automation of power systems. Related to the security analysis of the communication networks, a really powerful tool is the Stochastic Colored Petri Net (SCPN), because it models asynchronous and concurrent processes, and it is useful to analyze delays in timed systems. So, based in this context and knowing that time is really critical in transmitting messages in the IEC 61850 context; it is modeled in an IEC 61850 scenario a cyber-attack using SCPN that aims to delay the transmission of the messages, what is really critical for the operation of the system. The results corresponded to the modeled cyber-attack, showing the efficacy of the proposed modeling method.","PeriodicalId":285726,"journal":{"name":"2017 4th International Conference on Systems and Informatics (ICSAI)","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125811439","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 : 2017-11-01DOI: 10.1109/ICSAI.2017.8248279
Zhihao Wang, Q. Jia, Ping Ye, Hanxu Sun
The Visual SLAM algorithm has the goal of estimating the camera trajectory while reconstructing the Environment, which provides great help for autonomous navigation of mobile robot. However, many SLAM systems improve the complexity of the algorithm in order to show high precision, resulting in poor real-time performance. In practical application, considering the problem of robot life-time and the speed of embedded system, SLAM algorithm needs to improve the real-time performance of the algorithm and ensure the accuracy of application requirements at the same time. This paper presents a lightweight SLAM algorithm, which mainly through the management of map points, multi-threaded design and NEON technology to improve the real-tine performance of the algorithm while still maintaining a good localizing accuracy. It has great reference value for the practical application of SLAM algorithm.
{"title":"A depth camera based lightweight visual SLAM algorithm","authors":"Zhihao Wang, Q. Jia, Ping Ye, Hanxu Sun","doi":"10.1109/ICSAI.2017.8248279","DOIUrl":"https://doi.org/10.1109/ICSAI.2017.8248279","url":null,"abstract":"The Visual SLAM algorithm has the goal of estimating the camera trajectory while reconstructing the Environment, which provides great help for autonomous navigation of mobile robot. However, many SLAM systems improve the complexity of the algorithm in order to show high precision, resulting in poor real-time performance. In practical application, considering the problem of robot life-time and the speed of embedded system, SLAM algorithm needs to improve the real-time performance of the algorithm and ensure the accuracy of application requirements at the same time. This paper presents a lightweight SLAM algorithm, which mainly through the management of map points, multi-threaded design and NEON technology to improve the real-tine performance of the algorithm while still maintaining a good localizing accuracy. It has great reference value for the practical application of SLAM algorithm.","PeriodicalId":285726,"journal":{"name":"2017 4th International Conference on Systems and Informatics (ICSAI)","volume":"55 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126137299","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 : 2017-11-01DOI: 10.1109/ICSAI.2017.8248298
Min Tang, Linlin Ma, Bing Zhang, Cheng Wu, Liang Wang, E. Li, X. Chu
An optimal placement method for dynamic reactive power compensation devices is proposed in this paper to improve the immunity to commutation failures in multi-infeed HVDC systems. An electromagnetic transient simulation-based approach for computing the commutation failure immunity index is presented. An identification method is proposed for localizing zones in the receiving end power system where a fault will trigger simultaneous commutation failures of multiple HVDC infeeds. Various types of dynamic reactive power compensation devices are compared in their performance of improving commutation failure immunity. Finally, the optimal placement plan for dynamic reactive power compensation devices are to be made. Simulation results of New England 39-bus system verify the effectiveness of the proposed optimal placement method.
{"title":"Optimal placement of dynamic reactive power compensation devices for improving immunity to commutation failure in multi-infeed HVDC systems","authors":"Min Tang, Linlin Ma, Bing Zhang, Cheng Wu, Liang Wang, E. Li, X. Chu","doi":"10.1109/ICSAI.2017.8248298","DOIUrl":"https://doi.org/10.1109/ICSAI.2017.8248298","url":null,"abstract":"An optimal placement method for dynamic reactive power compensation devices is proposed in this paper to improve the immunity to commutation failures in multi-infeed HVDC systems. An electromagnetic transient simulation-based approach for computing the commutation failure immunity index is presented. An identification method is proposed for localizing zones in the receiving end power system where a fault will trigger simultaneous commutation failures of multiple HVDC infeeds. Various types of dynamic reactive power compensation devices are compared in their performance of improving commutation failure immunity. Finally, the optimal placement plan for dynamic reactive power compensation devices are to be made. Simulation results of New England 39-bus system verify the effectiveness of the proposed optimal placement method.","PeriodicalId":285726,"journal":{"name":"2017 4th International Conference on Systems and Informatics (ICSAI)","volume":"75 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123282244","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 : 2017-11-01DOI: 10.1109/ICSAI.2017.8248398
Yanxue Dong
The Image Classification Algorithm Based on Multi-feature Fusion and Deep Belief Networks is a method which extracts the color, texture and shape features of the image and integrates the three basic features first, and then, the fusion information is used as the input data of the deep belief networks model to train the samples and realize image classification. The results show that the classification accuracy can be improved by 21.2% compared with the image classification using a single feature. Compared with the mainstream classification algorithms, the classification accuracy can be effectively improved and it need no more time consuming.
{"title":"Image classification algorithm based on multi — Feature fusion and deep belief network","authors":"Yanxue Dong","doi":"10.1109/ICSAI.2017.8248398","DOIUrl":"https://doi.org/10.1109/ICSAI.2017.8248398","url":null,"abstract":"The Image Classification Algorithm Based on Multi-feature Fusion and Deep Belief Networks is a method which extracts the color, texture and shape features of the image and integrates the three basic features first, and then, the fusion information is used as the input data of the deep belief networks model to train the samples and realize image classification. The results show that the classification accuracy can be improved by 21.2% compared with the image classification using a single feature. Compared with the mainstream classification algorithms, the classification accuracy can be effectively improved and it need no more time consuming.","PeriodicalId":285726,"journal":{"name":"2017 4th International Conference on Systems and Informatics (ICSAI)","volume":"47 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115222656","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 : 2017-11-01DOI: 10.1109/ICSAI.2017.8248342
Kun Zhang, Dongping Zhang, Changxing Jing, Jianchao Li, Li Yang
Thanks to the recent development of deep convolutional neural networks, the performance of face verification has increased significantly. The softmax loss function is used mostly to make CNN models trained well. In order to get more robust face feature by deep CNN model, this paper proposes a new supervision signal based on the regular softmax loss function, namely scalable softmax loss, for face verification task. The scalable softmax loss function adjust the contribution to final loss for different training samples by a learned parameter. And, it's important to note that our proposed scalable softmax loss function can be easily implemented using existing deep learning frameworks. Extensive analysis and experiments on Labeled Face in the Wild(LFW) and YouTube Faces(YTF) show the superiority of the scalable softmax loss function in face verification task. Specially, our proposed scalable achieves comparable results on challenging LFW data set and YTF data set with the accuracy 99% and 95.08% respectively. Codes are released at1.
{"title":"Scalable softmax loss for face verification","authors":"Kun Zhang, Dongping Zhang, Changxing Jing, Jianchao Li, Li Yang","doi":"10.1109/ICSAI.2017.8248342","DOIUrl":"https://doi.org/10.1109/ICSAI.2017.8248342","url":null,"abstract":"Thanks to the recent development of deep convolutional neural networks, the performance of face verification has increased significantly. The softmax loss function is used mostly to make CNN models trained well. In order to get more robust face feature by deep CNN model, this paper proposes a new supervision signal based on the regular softmax loss function, namely scalable softmax loss, for face verification task. The scalable softmax loss function adjust the contribution to final loss for different training samples by a learned parameter. And, it's important to note that our proposed scalable softmax loss function can be easily implemented using existing deep learning frameworks. Extensive analysis and experiments on Labeled Face in the Wild(LFW) and YouTube Faces(YTF) show the superiority of the scalable softmax loss function in face verification task. Specially, our proposed scalable achieves comparable results on challenging LFW data set and YTF data set with the accuracy 99% and 95.08% respectively. Codes are released at1.","PeriodicalId":285726,"journal":{"name":"2017 4th International Conference on Systems and Informatics (ICSAI)","volume":"70 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114579484","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 : 2017-11-01DOI: 10.1109/ICSAI.2017.8248308
W. Chai, X. Cai, Zheng Li
The idea of applying energy storage system (ESS) into the demand side for accurate response has attracted lots of attention. Hybrid energy storage system (HESS) is the combination of various ESS types, which can exploit the strength of each ESS type and enjoy good comprehensive performance. In order to: 1) increase the accuracy of the demand side response, 2) extend the HESS lifetime, 3) reduce the energy loss of HESS and 4) lower the maintenance cost, a HESS synthetical model is developed including the lifetime model, the energy efficiency model, the economic cost model, the demand side response accuracy and the constraints. Comparison of different single types of ESS for demand side response is studied, and a multi-objective optimal control scheme of the HESS is proposed based on the particle swarm optimization (PSO). Finally, the effectiveness of the proposed control strategy is verified by actual data from a regional power grid in eastern China.
{"title":"A multi-objective optimal control scheme of the hybrid energy storage system for accurate response in the demand side","authors":"W. Chai, X. Cai, Zheng Li","doi":"10.1109/ICSAI.2017.8248308","DOIUrl":"https://doi.org/10.1109/ICSAI.2017.8248308","url":null,"abstract":"The idea of applying energy storage system (ESS) into the demand side for accurate response has attracted lots of attention. Hybrid energy storage system (HESS) is the combination of various ESS types, which can exploit the strength of each ESS type and enjoy good comprehensive performance. In order to: 1) increase the accuracy of the demand side response, 2) extend the HESS lifetime, 3) reduce the energy loss of HESS and 4) lower the maintenance cost, a HESS synthetical model is developed including the lifetime model, the energy efficiency model, the economic cost model, the demand side response accuracy and the constraints. Comparison of different single types of ESS for demand side response is studied, and a multi-objective optimal control scheme of the HESS is proposed based on the particle swarm optimization (PSO). Finally, the effectiveness of the proposed control strategy is verified by actual data from a regional power grid in eastern China.","PeriodicalId":285726,"journal":{"name":"2017 4th International Conference on Systems and Informatics (ICSAI)","volume":"99 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122041614","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：