Pub Date : 2021-12-10DOI: 10.1109/ICCSS53909.2021.9722013
Wang Hao, Nangfeng Xiao
The complex underwater environment and lighting conditions make underwater images suffer from texture distortion and color variations. In this paper, we propose an improved YOLOv4 detection method to detect four underwater organisms: holothurian, echinus, scallop, starfish and waterweeds. Firstly, we modified the network structure, added a deep separable convolution to the backbone network, and added a 152×152 feature map, which is conducive to the detection of small targets. Secondly, k-means clustering algorithm is used to cluster the bounding box in the data set, and the size of the bounding box is improved according to the clustering results. Thirdly, we propose a new module (EASPP, Spatial Pyramid Pooling), which increases slightly the model complexity, but the improvement effect is significant. Finally, when training the model, we use multi-scale training to better train targets with different scales. The experimental results show that on our test set, the improved method in the underwater object detection method is 4.8% higher than the original YOLOv4 model in accuracy (AP), the F1-score is 5.1% higher than that of the original method, and for mAP@0.5 it reaches 81.5%, which is 5.6% higher than that of the original method, which can be concluded that our method is effective.
{"title":"Research on Underwater Object Detection Based on Improved YOLOv4","authors":"Wang Hao, Nangfeng Xiao","doi":"10.1109/ICCSS53909.2021.9722013","DOIUrl":"https://doi.org/10.1109/ICCSS53909.2021.9722013","url":null,"abstract":"The complex underwater environment and lighting conditions make underwater images suffer from texture distortion and color variations. In this paper, we propose an improved YOLOv4 detection method to detect four underwater organisms: holothurian, echinus, scallop, starfish and waterweeds. Firstly, we modified the network structure, added a deep separable convolution to the backbone network, and added a 152×152 feature map, which is conducive to the detection of small targets. Secondly, k-means clustering algorithm is used to cluster the bounding box in the data set, and the size of the bounding box is improved according to the clustering results. Thirdly, we propose a new module (EASPP, Spatial Pyramid Pooling), which increases slightly the model complexity, but the improvement effect is significant. Finally, when training the model, we use multi-scale training to better train targets with different scales. The experimental results show that on our test set, the improved method in the underwater object detection method is 4.8% higher than the original YOLOv4 model in accuracy (AP), the F1-score is 5.1% higher than that of the original method, and for mAP@0.5 it reaches 81.5%, which is 5.6% higher than that of the original method, which can be concluded that our method is effective.","PeriodicalId":435816,"journal":{"name":"2021 8th International Conference on Information, Cybernetics, and Computational Social Systems (ICCSS)","volume":"42 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132983085","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 : 2021-12-10DOI: 10.1109/ICCSS53909.2021.9722036
Liu Yong, Yingying Chi, Zheng Zhe, Liu Rui, Cui Wenpeng, Jia Xiaoguang
Based on summary of the main international standards of Ethernet physical layer chip and its working principle, this paper analyzed the requirements of the Ethernet physical layer chip in different Internet of energy terminals of Smart Substation, Distribution communication, Smart power plant, such as the communication bandwidth, electromagnetic compatibility, operating temperature. Technical requirements such as time-synchronization and energy saving also are summarized. The Ethernet physical layer core supporting IEEE1588 timestamp function can improve the sampling synchronization accuracy, and the physical layer chip supporting IEEE802.3az high efficiency and energy saving Ethernet standard can significantly reduce the energy consumption and cost of the device for power distribution devices powered by off-line power supply.
{"title":"Overview of Ethernet Physical Layer Chip Technology for Internet of energy Terminal","authors":"Liu Yong, Yingying Chi, Zheng Zhe, Liu Rui, Cui Wenpeng, Jia Xiaoguang","doi":"10.1109/ICCSS53909.2021.9722036","DOIUrl":"https://doi.org/10.1109/ICCSS53909.2021.9722036","url":null,"abstract":"Based on summary of the main international standards of Ethernet physical layer chip and its working principle, this paper analyzed the requirements of the Ethernet physical layer chip in different Internet of energy terminals of Smart Substation, Distribution communication, Smart power plant, such as the communication bandwidth, electromagnetic compatibility, operating temperature. Technical requirements such as time-synchronization and energy saving also are summarized. The Ethernet physical layer core supporting IEEE1588 timestamp function can improve the sampling synchronization accuracy, and the physical layer chip supporting IEEE802.3az high efficiency and energy saving Ethernet standard can significantly reduce the energy consumption and cost of the device for power distribution devices powered by off-line power supply.","PeriodicalId":435816,"journal":{"name":"2021 8th International Conference on Information, Cybernetics, and Computational Social Systems (ICCSS)","volume":"75 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132673771","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 : 2021-12-10DOI: 10.1109/ICCSS53909.2021.9721954
Yifa Liu, Wenchao Xue, S. He, Long Cheng
The power grid is a kind of national critical infrastructure directly affiliated to human daily life. Because of the vital functions and potentially significant losses, the power grid becomes an excellent target for many malicious attacks. Due to the special nonlinear measurements, many detection methods do not match the grid very well. The extended Kalman filter based detection is one of the few methods suitable for nonlinear system detection, and therefore can be used in power system to spot malicious attacks. However, the reliability and effectiveness of the extended Kalman filter based detection have not been fully studied and adequately guaranteed. By proposing a two-step false data injection attack strategy, this paper gives a stealthy way to inject false data of increasing magnitude into the power grid, which can eventually cause a certain degree of deviation of the grid state without being detected. In the simulation, the method proposed in this paper caused a voltage deviation of more than 25% before being discovered in the power system.
{"title":"Stealthy False Data Injection Attacks against Extended Kalman Filter Detection in Power Grids","authors":"Yifa Liu, Wenchao Xue, S. He, Long Cheng","doi":"10.1109/ICCSS53909.2021.9721954","DOIUrl":"https://doi.org/10.1109/ICCSS53909.2021.9721954","url":null,"abstract":"The power grid is a kind of national critical infrastructure directly affiliated to human daily life. Because of the vital functions and potentially significant losses, the power grid becomes an excellent target for many malicious attacks. Due to the special nonlinear measurements, many detection methods do not match the grid very well. The extended Kalman filter based detection is one of the few methods suitable for nonlinear system detection, and therefore can be used in power system to spot malicious attacks. However, the reliability and effectiveness of the extended Kalman filter based detection have not been fully studied and adequately guaranteed. By proposing a two-step false data injection attack strategy, this paper gives a stealthy way to inject false data of increasing magnitude into the power grid, which can eventually cause a certain degree of deviation of the grid state without being detected. In the simulation, the method proposed in this paper caused a voltage deviation of more than 25% before being discovered in the power system.","PeriodicalId":435816,"journal":{"name":"2021 8th International Conference on Information, Cybernetics, and Computational Social Systems (ICCSS)","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116424628","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 : 2021-12-10DOI: 10.1109/ICCSS53909.2021.9722005
Xiaoye Fan, Xiaolong Wu, Hong-gui Han
The feature extraction method is an effective tool to understand the behavior of plug-flow wastewater treatment process (PF-WWTP). However, it is a challenge to extract feature components due to PF-WWTP subjected to the time-varying system with dataset mismatch. To solve this problem, in this paper, an adaptive feature extraction method (AFEM) based on dynamic kernel principal component analysis (KPCA) is proposed to improve the feature extraction accuracy. First, a data adjustment method is proposed to adapt datasets of process variables to the different hydraulic residence time. Then, the matching datasets can be used to observe the dynamics of metabolism within PF-WWTP. Second, a dynamic KPCA algorithm based on iterative calculation is introduced to obtain the contribution of feature components for process variables. This algorithm can update the order of feature components online following with the time-varying flow-rates of PF-WWTP. Third, an error-oriented self-adaptive mechanism is designed to determine the dimension of feature components for process variables. This mechanism not only performs preferable feature extraction without giving thresholds but also ensures its realtime accuracy. Finally, AFEM is compared with some existing feature extraction methods through experiments. The results show that the proposed AFEM can accurately extract feature components for PF-WWTP.
{"title":"Dynamic KPCA for Feature Extraction of Wastewater Treatment Process","authors":"Xiaoye Fan, Xiaolong Wu, Hong-gui Han","doi":"10.1109/ICCSS53909.2021.9722005","DOIUrl":"https://doi.org/10.1109/ICCSS53909.2021.9722005","url":null,"abstract":"The feature extraction method is an effective tool to understand the behavior of plug-flow wastewater treatment process (PF-WWTP). However, it is a challenge to extract feature components due to PF-WWTP subjected to the time-varying system with dataset mismatch. To solve this problem, in this paper, an adaptive feature extraction method (AFEM) based on dynamic kernel principal component analysis (KPCA) is proposed to improve the feature extraction accuracy. First, a data adjustment method is proposed to adapt datasets of process variables to the different hydraulic residence time. Then, the matching datasets can be used to observe the dynamics of metabolism within PF-WWTP. Second, a dynamic KPCA algorithm based on iterative calculation is introduced to obtain the contribution of feature components for process variables. This algorithm can update the order of feature components online following with the time-varying flow-rates of PF-WWTP. Third, an error-oriented self-adaptive mechanism is designed to determine the dimension of feature components for process variables. This mechanism not only performs preferable feature extraction without giving thresholds but also ensures its realtime accuracy. Finally, AFEM is compared with some existing feature extraction methods through experiments. The results show that the proposed AFEM can accurately extract feature components for PF-WWTP.","PeriodicalId":435816,"journal":{"name":"2021 8th International Conference on Information, Cybernetics, and Computational Social Systems (ICCSS)","volume":"32 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123974089","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 : 2021-12-10DOI: 10.1109/ICCSS53909.2021.9722015
Xuejing Lan, Weijie Yang, Jianing Zhang, Zhijia Zhao, Ge Ma, Zhifu Li
This paper considers the decoupling control problem of a two-degree-of-freedom (2-DOF) helicopter system with uncertainties and disturbances. The unknown input bias caused by the dynamical coupling is approximated by fuzzy neural networks. An adaptive sliding mode control (SMC) strategy is proposed to deal with the uncertainties and unknown disturbances on the system. By appropriately constructing the Lyapunov function, the stability of the controlled system is proved. Finally, the effectiveness and availability of the strategy are verified by numerical simulation.
{"title":"Sliding mode control of a 2-DOF helicopter system with adaptive input compensation","authors":"Xuejing Lan, Weijie Yang, Jianing Zhang, Zhijia Zhao, Ge Ma, Zhifu Li","doi":"10.1109/ICCSS53909.2021.9722015","DOIUrl":"https://doi.org/10.1109/ICCSS53909.2021.9722015","url":null,"abstract":"This paper considers the decoupling control problem of a two-degree-of-freedom (2-DOF) helicopter system with uncertainties and disturbances. The unknown input bias caused by the dynamical coupling is approximated by fuzzy neural networks. An adaptive sliding mode control (SMC) strategy is proposed to deal with the uncertainties and unknown disturbances on the system. By appropriately constructing the Lyapunov function, the stability of the controlled system is proved. Finally, the effectiveness and availability of the strategy are verified by numerical simulation.","PeriodicalId":435816,"journal":{"name":"2021 8th International Conference on Information, Cybernetics, and Computational Social Systems (ICCSS)","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128168951","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}
Neural Architecture Search (NAS) is proposed to automatically search novel neural networks. Currently, one typical problem of NAS is that its computation requirements are too high to stand for most researchers. In fact, it consumes a lot of resources to train subnetworks for architecture search. If the performance of each subnetwork can be predicted accurately without training, the computational burden will be alleviated. Graph Convolutional Network (GCN) is proven to have powerful capabilities for topological information perception and extraction. It is suitable to use GCN for predicting neural architecture performance which is related to its topology.In this paper, we treat GCN as the performance predictor with two improvements. First, a novel neural architecture data processing method named DATAPRO2 is designed to improve GCN’s performance. Then, we propose EGCN, a model-based performance predictor which employs ensemble technique on GCN with DATAPRO2 to alleviate the overfitting issue caused by the imbalanced dataset for neural architecture performance prediction. Experimental results on CVPR-2021-NAS-TRACK2 dataset show that EGCN contributes to obtaining better predictive performance than vanilla GCN and other popular predictors.
{"title":"EGCN: Ensemble Graph Convolutional Network for Neural Architecture Performance Prediction","authors":"Xin Liu, Zixiang Ding, Nannan Li, Yaran Chen, Dong Zhao","doi":"10.1109/ICCSS53909.2021.9721976","DOIUrl":"https://doi.org/10.1109/ICCSS53909.2021.9721976","url":null,"abstract":"Neural Architecture Search (NAS) is proposed to automatically search novel neural networks. Currently, one typical problem of NAS is that its computation requirements are too high to stand for most researchers. In fact, it consumes a lot of resources to train subnetworks for architecture search. If the performance of each subnetwork can be predicted accurately without training, the computational burden will be alleviated. Graph Convolutional Network (GCN) is proven to have powerful capabilities for topological information perception and extraction. It is suitable to use GCN for predicting neural architecture performance which is related to its topology.In this paper, we treat GCN as the performance predictor with two improvements. First, a novel neural architecture data processing method named DATAPRO2 is designed to improve GCN’s performance. Then, we propose EGCN, a model-based performance predictor which employs ensemble technique on GCN with DATAPRO2 to alleviate the overfitting issue caused by the imbalanced dataset for neural architecture performance prediction. Experimental results on CVPR-2021-NAS-TRACK2 dataset show that EGCN contributes to obtaining better predictive performance than vanilla GCN and other popular predictors.","PeriodicalId":435816,"journal":{"name":"2021 8th International Conference on Information, Cybernetics, and Computational Social Systems (ICCSS)","volume":"463 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129583783","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 : 2021-12-10DOI: 10.1109/ICCSS53909.2021.9721990
Haoyu Li, Leimin Wang
This paper investigates a unified controller to solve the fixed-time anti-synchronization (FTAS) and preassigned-time anti-synchronization (PTAS) problems for chaotic neural networks. Under our controller, chaotic neural network can realize anti-synchronization within the fixed or preassigned time which greatly expands the practical application range of the anti-synchronization. In addition, sufficient conditions and time estimation on FTAS and PTAS are derived. Finally, the feasibility of the control scheme is proved via a numerical simulation.
{"title":"Fixed-/Preassigned-Time Anti-Synchronization of Chaotic Neural Networks","authors":"Haoyu Li, Leimin Wang","doi":"10.1109/ICCSS53909.2021.9721990","DOIUrl":"https://doi.org/10.1109/ICCSS53909.2021.9721990","url":null,"abstract":"This paper investigates a unified controller to solve the fixed-time anti-synchronization (FTAS) and preassigned-time anti-synchronization (PTAS) problems for chaotic neural networks. Under our controller, chaotic neural network can realize anti-synchronization within the fixed or preassigned time which greatly expands the practical application range of the anti-synchronization. In addition, sufficient conditions and time estimation on FTAS and PTAS are derived. Finally, the feasibility of the control scheme is proved via a numerical simulation.","PeriodicalId":435816,"journal":{"name":"2021 8th International Conference on Information, Cybernetics, and Computational Social Systems (ICCSS)","volume":"34 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129077818","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}
Automatic crane is a complex system affected by the external environment and the internal components of the system, information fusion, software and hardware combination, and man-machine integration. The improvement of its automation and informatization proposes various challenges in the accident model construction and safety analysis. However, the safety analysis methods based on fault types consider that the occurrence of accidents is linear and ignore the correlation among components of the system. This paper adopts the system-theoretic accident model and process (STAMP) and system-theoretic process analysis (STPA) mode is to implement safety analysis of the automatic crane trolley running system (ACTRs). The paper starts from the identification of system-level losses and hazards, clarifies the function and internal logical control relationships of the system’s components, and then finds potential unsafe control actions (UCAs) and loss scenarios during the trolley running. The results show that the control requirements for the regular operation of the trolley running system can be analyzed in detail. Therefore, the STAMP/STPA can apply to the safety investigation of automatic cranes.
{"title":"Safety Analysis of Automatic Crane Trolley Running System Based on STAMP/STPA","authors":"Wenbo Zhang, Xiangkun Meng, Jianyuan Wang, Tie-shan Li, Qihe Shan, Fei Teng","doi":"10.1109/ICCSS53909.2021.9722016","DOIUrl":"https://doi.org/10.1109/ICCSS53909.2021.9722016","url":null,"abstract":"Automatic crane is a complex system affected by the external environment and the internal components of the system, information fusion, software and hardware combination, and man-machine integration. The improvement of its automation and informatization proposes various challenges in the accident model construction and safety analysis. However, the safety analysis methods based on fault types consider that the occurrence of accidents is linear and ignore the correlation among components of the system. This paper adopts the system-theoretic accident model and process (STAMP) and system-theoretic process analysis (STPA) mode is to implement safety analysis of the automatic crane trolley running system (ACTRs). The paper starts from the identification of system-level losses and hazards, clarifies the function and internal logical control relationships of the system’s components, and then finds potential unsafe control actions (UCAs) and loss scenarios during the trolley running. The results show that the control requirements for the regular operation of the trolley running system can be analyzed in detail. Therefore, the STAMP/STPA can apply to the safety investigation of automatic cranes.","PeriodicalId":435816,"journal":{"name":"2021 8th International Conference on Information, Cybernetics, and Computational Social Systems (ICCSS)","volume":"46 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115264368","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 : 2021-12-10DOI: 10.1109/ICCSS53909.2021.9721944
Haixiao Zhao, Rongrong Wang, Jin Zhou, Shiyuan Han, Tao Du, Ke Ji, Ya-ou Zhao, Kun Zhang, Yuehui Chen
As an effective deep clustering method, improved deep embedding clustering can process large-scale high-dimensional data. However, the method only focuses on the global data and does not consider the local graph structure between data points. In this paper, a semi-supervised deep clustering algorithm with soft membership affinity is proposed to cluster high-dimensional datasets. The proposed algorithm is composed of three parts: the reconstruction loss is adopted to recover data and extract important features on latent space, the KL divergence between the soft assignment and the target distribution is utilized to make samples in each cluster distribute more densely, and the novel soft membership affinity, which is regarded as the semi-supervised information, is introduced to the IDEC model to constrain the relationship between data points and their neighbors, so as to further enhance the clustering performance. Experiments on datasets show that the algorithm is effective compared with other deep clustering algorithms.
{"title":"Semi-Supervised Deep Clustering with Soft Membership Affinity","authors":"Haixiao Zhao, Rongrong Wang, Jin Zhou, Shiyuan Han, Tao Du, Ke Ji, Ya-ou Zhao, Kun Zhang, Yuehui Chen","doi":"10.1109/ICCSS53909.2021.9721944","DOIUrl":"https://doi.org/10.1109/ICCSS53909.2021.9721944","url":null,"abstract":"As an effective deep clustering method, improved deep embedding clustering can process large-scale high-dimensional data. However, the method only focuses on the global data and does not consider the local graph structure between data points. In this paper, a semi-supervised deep clustering algorithm with soft membership affinity is proposed to cluster high-dimensional datasets. The proposed algorithm is composed of three parts: the reconstruction loss is adopted to recover data and extract important features on latent space, the KL divergence between the soft assignment and the target distribution is utilized to make samples in each cluster distribute more densely, and the novel soft membership affinity, which is regarded as the semi-supervised information, is introduced to the IDEC model to constrain the relationship between data points and their neighbors, so as to further enhance the clustering performance. Experiments on datasets show that the algorithm is effective compared with other deep clustering algorithms.","PeriodicalId":435816,"journal":{"name":"2021 8th International Conference on Information, Cybernetics, and Computational Social Systems (ICCSS)","volume":"16 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114035129","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 : 2021-12-10DOI: 10.1109/ICCSS53909.2021.9722026
Kaihan Yu, Degang Wang, Hongxing Li
In this paper, a prediction model based on a broad learning system (BLS) and temporal convolutional network (TCN) is proposed to measure the remaining useful life (RUL) of turbofan engines. Firstly, a variational autoencoder (VAE) is used to extract important low-dimensional features from the engine sensor data. Then, the degradation information is extracted from the time and feature dimensions of fragment data using TCN. Further, the BLS combined with residual connection is used to enhance the nonlinear representation of the model. The proposed method is validated on the commercial modular aero propulsion system simulation (C-MAPSS) dataset and compared with some state-of-the-art methods. The experimental results show that the proposed method is effective in RUL prediction.
{"title":"A Prediction Model for Remaining Useful Life of Turbofan Engines by Fusing Broad Learning System and Temporal Convolutional Network","authors":"Kaihan Yu, Degang Wang, Hongxing Li","doi":"10.1109/ICCSS53909.2021.9722026","DOIUrl":"https://doi.org/10.1109/ICCSS53909.2021.9722026","url":null,"abstract":"In this paper, a prediction model based on a broad learning system (BLS) and temporal convolutional network (TCN) is proposed to measure the remaining useful life (RUL) of turbofan engines. Firstly, a variational autoencoder (VAE) is used to extract important low-dimensional features from the engine sensor data. Then, the degradation information is extracted from the time and feature dimensions of fragment data using TCN. Further, the BLS combined with residual connection is used to enhance the nonlinear representation of the model. The proposed method is validated on the commercial modular aero propulsion system simulation (C-MAPSS) dataset and compared with some state-of-the-art methods. The experimental results show that the proposed method is effective in RUL prediction.","PeriodicalId":435816,"journal":{"name":"2021 8th International Conference on Information, Cybernetics, and Computational Social Systems (ICCSS)","volume":"47 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122865961","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：