Pub Date : 2022-07-22DOI: 10.1109/ISPDS56360.2022.9874015
Tian-Fei Zhang, J. Ding, Rong-Qiang Zhou, Haiyan Long
Aiming at the low accuracy of crowd counting caused by scale change and occlusion in dense scenes, this paper proposes to generate the truth map into non overlapping independent areas in HRNet to facilitate the crowd location statistics of network density map; Then the 3D attention mechanism is introduced to make the network focus on the useful information of the feature map; Finally, during the training, the mean square error loss (MSE loss), L1 loss and cross entropy loss are combined into the total loss function to optimize the generalization ability of the model; The combination of the above methods improves the accuracy of the model in crowd counting and crowd location. Compared with the main methods in recent years in the public datasets NWPU, Shanghai Tech, the experimental results show that the proposed model can effectively improve the accuracy and robustness of crowd location counting.
{"title":"An Improved HRNET and its application in crowd counting","authors":"Tian-Fei Zhang, J. Ding, Rong-Qiang Zhou, Haiyan Long","doi":"10.1109/ISPDS56360.2022.9874015","DOIUrl":"https://doi.org/10.1109/ISPDS56360.2022.9874015","url":null,"abstract":"Aiming at the low accuracy of crowd counting caused by scale change and occlusion in dense scenes, this paper proposes to generate the truth map into non overlapping independent areas in HRNet to facilitate the crowd location statistics of network density map; Then the 3D attention mechanism is introduced to make the network focus on the useful information of the feature map; Finally, during the training, the mean square error loss (MSE loss), L1 loss and cross entropy loss are combined into the total loss function to optimize the generalization ability of the model; The combination of the above methods improves the accuracy of the model in crowd counting and crowd location. Compared with the main methods in recent years in the public datasets NWPU, Shanghai Tech, the experimental results show that the proposed model can effectively improve the accuracy and robustness of crowd location counting.","PeriodicalId":280244,"journal":{"name":"2022 3rd International Conference on Information Science, Parallel and Distributed Systems (ISPDS)","volume":"56 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116573391","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}
Hyperspectral images (HSIs) will experience noise throughout the data collection process due to the imaging system's limitations, which will make it challenging to extract the image's crucial information. In this paper, a multi-stage enhanced HSI denoising network (MED-Net) is proposed. Our core concept is to process the hyperspectral noise image iteratively using a multi-stage network. A similar network structure's first and second phases are employed for the denoise process. To achieve cross-stage information transfer, we use CSFF (Cross-stage Feature Fusion) mechanism and SAM (Supervised Attention Module). AN (Additive Network) and MN (Multiplicative Network) are used to remove additive and multiplicative noise. Then, we restore the background based on the residual network and attention mechanism. The results of our experiments demonstrate the superiority of our approach over the actual HSIs data recovery, and the restored image has good visual clarity and detail.
{"title":"Multi-stage Enhanced Denoising Network on Hyperspectral Image","authors":"Xiaomiao Pan, Q. Pan, Chao Wang, Chuan-Sheng Yang, Yueting Yang, Liangtian He","doi":"10.1109/ISPDS56360.2022.9874188","DOIUrl":"https://doi.org/10.1109/ISPDS56360.2022.9874188","url":null,"abstract":"Hyperspectral images (HSIs) will experience noise throughout the data collection process due to the imaging system's limitations, which will make it challenging to extract the image's crucial information. In this paper, a multi-stage enhanced HSI denoising network (MED-Net) is proposed. Our core concept is to process the hyperspectral noise image iteratively using a multi-stage network. A similar network structure's first and second phases are employed for the denoise process. To achieve cross-stage information transfer, we use CSFF (Cross-stage Feature Fusion) mechanism and SAM (Supervised Attention Module). AN (Additive Network) and MN (Multiplicative Network) are used to remove additive and multiplicative noise. Then, we restore the background based on the residual network and attention mechanism. The results of our experiments demonstrate the superiority of our approach over the actual HSIs data recovery, and the restored image has good visual clarity and detail.","PeriodicalId":280244,"journal":{"name":"2022 3rd International Conference on Information Science, Parallel and Distributed Systems (ISPDS)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126075427","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 : 2022-07-22DOI: 10.1109/ISPDS56360.2022.9874107
Xianguo Wang, Chunxi Guan, Shunjie Lin, Hanhua Cao
In biometric identification, the most widely used form was the fingerprint, which has unique and invariable property. The main task of fingerprint enhancement was to restore the structural defects of its ridge. The ultimate goal was to improve the accuracy of fingerprint feature extraction by improving the quality of the ridge, leading to improving the accuracy of fingerprint identification. Based on the fingerprint enhancement algorithms, The Gabor filtering has nice band pass ability in fingerprint enhancement from experimental results, with nice direction and frequency selectivity. Thus, Gabor filtering can effectively remove the noise of the ridge along its direction, also save the true ridge structure. As an improved algorithm, Log-Gabor filter which can make up the defect of Gabor filter improves the final effect of filtering.
{"title":"Research on Fingerprint Image Based on Improved Log-Gabor Filter Enhancement Algorithm","authors":"Xianguo Wang, Chunxi Guan, Shunjie Lin, Hanhua Cao","doi":"10.1109/ISPDS56360.2022.9874107","DOIUrl":"https://doi.org/10.1109/ISPDS56360.2022.9874107","url":null,"abstract":"In biometric identification, the most widely used form was the fingerprint, which has unique and invariable property. The main task of fingerprint enhancement was to restore the structural defects of its ridge. The ultimate goal was to improve the accuracy of fingerprint feature extraction by improving the quality of the ridge, leading to improving the accuracy of fingerprint identification. Based on the fingerprint enhancement algorithms, The Gabor filtering has nice band pass ability in fingerprint enhancement from experimental results, with nice direction and frequency selectivity. Thus, Gabor filtering can effectively remove the noise of the ridge along its direction, also save the true ridge structure. As an improved algorithm, Log-Gabor filter which can make up the defect of Gabor filter improves the final effect of filtering.","PeriodicalId":280244,"journal":{"name":"2022 3rd International Conference on Information Science, Parallel and Distributed Systems (ISPDS)","volume":"35 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121753681","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 : 2022-07-22DOI: 10.1109/ISPDS56360.2022.9874134
Nan Yang, Chunlin He
Aiming at the low accuracy and time-consuming training of malaria detection, this paper proposes a malaria detection algorithm based on ResNet+CBAM attention mechanism. In the ResNet-40 model, which reduces the number of network layers and network width, the CBAM attention mechanism module is added and trained on the malaria dataset (Malaria dataset). The experimental results show that the detection method proposed in this paper improves the classification accuracy by 1% on the original basis.
{"title":"Malaria detection based on ResNet + CBAM attention mechanism","authors":"Nan Yang, Chunlin He","doi":"10.1109/ISPDS56360.2022.9874134","DOIUrl":"https://doi.org/10.1109/ISPDS56360.2022.9874134","url":null,"abstract":"Aiming at the low accuracy and time-consuming training of malaria detection, this paper proposes a malaria detection algorithm based on ResNet+CBAM attention mechanism. In the ResNet-40 model, which reduces the number of network layers and network width, the CBAM attention mechanism module is added and trained on the malaria dataset (Malaria dataset). The experimental results show that the detection method proposed in this paper improves the classification accuracy by 1% on the original basis.","PeriodicalId":280244,"journal":{"name":"2022 3rd International Conference on Information Science, Parallel and Distributed Systems (ISPDS)","volume":"212 5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132618456","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 : 2022-07-22DOI: 10.1109/ISPDS56360.2022.9874250
Zhiying Yu, A. Li, Fan Yu
Aiming at the problem of cultivating applied talents to serve the society, the electronic system comprehensive design experiment takes MCU as the core, and designs an intelligent trash can. The trash can can automatically control the opening and closing of the trash can lid according to the infrared sensor to detect whether someone puts garbage; At the same time, it has the function of manual one-key control of opening and closing the trash can lid and the overflow sound and light alarm function; The trash can has GPS positioning, GPRS wireless communication function, the garbage processing center can query the status and geographic location information of the trash can in real time through the Internet of Things platform. This pro-ject is closely related to the practical application of life and inte-grates multiple knowledge points. The student-centered experi-mental teaching mode improves students' autonomy in learning, stimulates students' creativity, and improves students' ability to solve practical problems. The teaching effect is good.
{"title":"Electronic System Comprehensive Design Experiment-Intelligent Trash Can Design","authors":"Zhiying Yu, A. Li, Fan Yu","doi":"10.1109/ISPDS56360.2022.9874250","DOIUrl":"https://doi.org/10.1109/ISPDS56360.2022.9874250","url":null,"abstract":"Aiming at the problem of cultivating applied talents to serve the society, the electronic system comprehensive design experiment takes MCU as the core, and designs an intelligent trash can. The trash can can automatically control the opening and closing of the trash can lid according to the infrared sensor to detect whether someone puts garbage; At the same time, it has the function of manual one-key control of opening and closing the trash can lid and the overflow sound and light alarm function; The trash can has GPS positioning, GPRS wireless communication function, the garbage processing center can query the status and geographic location information of the trash can in real time through the Internet of Things platform. This pro-ject is closely related to the practical application of life and inte-grates multiple knowledge points. The student-centered experi-mental teaching mode improves students' autonomy in learning, stimulates students' creativity, and improves students' ability to solve practical problems. The teaching effect is good.","PeriodicalId":280244,"journal":{"name":"2022 3rd International Conference on Information Science, Parallel and Distributed Systems (ISPDS)","volume":"3 2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124204213","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 : 2022-07-22DOI: 10.1109/ISPDS56360.2022.9874251
Feifan Gu, Zhaohui Meng
Video super resolution technology refers to the reconstruction of low resolution video into high resolution frames. In recent years, the application of deep learning to super-resolution technology has attracted extensive attention. However, the reconstruction effect of the existing model still has some problems such as double shadow, structural loss, and the solutions of problems are relatively rare. In this paper, we propose a new idea to use gradient extraction branches to guide the reconstruction of high resolution frames in backbone networks. The loss function is improved by combining gradient loss with pixel loss to improve convergence ability. Multi-scale convolution is introduced into the alignment module to enlarge the receptive field and improve the performance of the model to extract large motion features. Experimental results show that the model has good performance on REDS4 and Vid4 data sets.
{"title":"Structure-Preserving Video Super Resolution with Multi-Scale Convolution","authors":"Feifan Gu, Zhaohui Meng","doi":"10.1109/ISPDS56360.2022.9874251","DOIUrl":"https://doi.org/10.1109/ISPDS56360.2022.9874251","url":null,"abstract":"Video super resolution technology refers to the reconstruction of low resolution video into high resolution frames. In recent years, the application of deep learning to super-resolution technology has attracted extensive attention. However, the reconstruction effect of the existing model still has some problems such as double shadow, structural loss, and the solutions of problems are relatively rare. In this paper, we propose a new idea to use gradient extraction branches to guide the reconstruction of high resolution frames in backbone networks. The loss function is improved by combining gradient loss with pixel loss to improve convergence ability. Multi-scale convolution is introduced into the alignment module to enlarge the receptive field and improve the performance of the model to extract large motion features. Experimental results show that the model has good performance on REDS4 and Vid4 data sets.","PeriodicalId":280244,"journal":{"name":"2022 3rd International Conference on Information Science, Parallel and Distributed Systems (ISPDS)","volume":"48 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124318402","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 : 2022-07-22DOI: 10.1109/ISPDS56360.2022.9874223
Chenzhao Huang, Mingrui Ji, Hang Zhang, Ruisen Luo
The previous modulation recognition models based on deep learning ignore the signal's complex characteristics and only consider the information carried by the signal in a single dimension, resulting in poor performance. Aiming at the complex characteristics of in-phase/quadrature (I/Q) data, this paper adopts a combination of complex convolution and one-dimensional real convolution, emphasizing the feature interaction between I and Q and enriching the feature representation of the signal. Besides, a multi-level complex attention block is introduced to enhance the informative representation of the entire feature space. Experimental results indicate that the proposed method's recognition accuracy of MQAM is significantly improved. Furthermore, the proposed method also alleviates the poor performance under a low signal-to-noise ratio, which is overall better than other deep learning-based modulation recognition models.
{"title":"A Multi-level Complex Feature Mining Method Based on Deep Learning for Automatic Modulation Recognition","authors":"Chenzhao Huang, Mingrui Ji, Hang Zhang, Ruisen Luo","doi":"10.1109/ISPDS56360.2022.9874223","DOIUrl":"https://doi.org/10.1109/ISPDS56360.2022.9874223","url":null,"abstract":"The previous modulation recognition models based on deep learning ignore the signal's complex characteristics and only consider the information carried by the signal in a single dimension, resulting in poor performance. Aiming at the complex characteristics of in-phase/quadrature (I/Q) data, this paper adopts a combination of complex convolution and one-dimensional real convolution, emphasizing the feature interaction between I and Q and enriching the feature representation of the signal. Besides, a multi-level complex attention block is introduced to enhance the informative representation of the entire feature space. Experimental results indicate that the proposed method's recognition accuracy of MQAM is significantly improved. Furthermore, the proposed method also alleviates the poor performance under a low signal-to-noise ratio, which is overall better than other deep learning-based modulation recognition models.","PeriodicalId":280244,"journal":{"name":"2022 3rd International Conference on Information Science, Parallel and Distributed Systems (ISPDS)","volume":"37 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116912239","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 : 2022-07-22DOI: 10.1109/ISPDS56360.2022.9874181
Yin Zhang, Xin Sun
Concerning the fact that the real-time video communication of mobile terminals is prone to video freeze due to the mobility of public network digital trunking terminal, a video bitrate adaptive algorithm for public network digital trunking terminals is proposed. Firstly, the mobile terminal detects the current geographic location and bandwidth situation in real time. Secondly, the prediction algorithm based on unscented Kalman filtering is used to predict the geographic location of the mobile terminal at the next moment. At the same time, the bandwidth situation corresponding to the geographic location at the next moment in the database is searched. Finally, dynamically adjust the video bit rate according to the current bandwidth situation and the bandwidth situation of the geographic location at the next moment. The experimental results show that, the proposed algorithm can accurately predict the geographic location at the next moment, and the geographic location prediction error is 1.89 meters; Compared with the video bitrate adaptation algorithm based on the current bandwidth information, the proposed algorithm can effectively reduce the video freeze and improve the video quality.
{"title":"A Video Bitrate Adaptive Algorithm for Public Network Digital Trunking Terminals","authors":"Yin Zhang, Xin Sun","doi":"10.1109/ISPDS56360.2022.9874181","DOIUrl":"https://doi.org/10.1109/ISPDS56360.2022.9874181","url":null,"abstract":"Concerning the fact that the real-time video communication of mobile terminals is prone to video freeze due to the mobility of public network digital trunking terminal, a video bitrate adaptive algorithm for public network digital trunking terminals is proposed. Firstly, the mobile terminal detects the current geographic location and bandwidth situation in real time. Secondly, the prediction algorithm based on unscented Kalman filtering is used to predict the geographic location of the mobile terminal at the next moment. At the same time, the bandwidth situation corresponding to the geographic location at the next moment in the database is searched. Finally, dynamically adjust the video bit rate according to the current bandwidth situation and the bandwidth situation of the geographic location at the next moment. The experimental results show that, the proposed algorithm can accurately predict the geographic location at the next moment, and the geographic location prediction error is 1.89 meters; Compared with the video bitrate adaptation algorithm based on the current bandwidth information, the proposed algorithm can effectively reduce the video freeze and improve the video quality.","PeriodicalId":280244,"journal":{"name":"2022 3rd International Conference on Information Science, Parallel and Distributed Systems (ISPDS)","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115682181","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 : 2022-07-22DOI: 10.1109/ISPDS56360.2022.9874146
Qianqiu Wang, Junjie Li, Xianlu Luo, Chun Chen
Due to the limitations imposed by and complexity of indoor environments, a low-cost and accurate indoor positioning system has not yet been designed. To address this issue, we constructed a fused indoor positioning algorithm based on the extended Kalman filter for WiFi and inertial measurement units (IMUs) using only a smartphone. To reduce the influence of WiFi signal fluctuation on fingerprint-based positioning, we used Gaussian process regression for denoising the data. We used our proposed improved clustering algorithm to reduce the matching amount in the positioning stage and increase the positioning accuracy. In terms of pedestrian dead reckoning (PDR) positioning, we designed a new and effective direction estimation algorithm integrating accelerometer and magnetometer, and we used an online step size estimation model to improve the accuracy of step size estimation. The experimental results showed that the average positioning error of the proposed fusion algorithm is 1.76 m, which was 55% lower than that using the WiFi network only, and 62% lower than using PDR only. Our findings showed that the fused positioning scheme based on WiFi and IMU can be used to effectively increase indoor positioning accuracy, and the proposed system is suitable for high-precision positioning scenarios.
{"title":"Fusion Algorithm of WiFi and IMU for Indoor Positioning","authors":"Qianqiu Wang, Junjie Li, Xianlu Luo, Chun Chen","doi":"10.1109/ISPDS56360.2022.9874146","DOIUrl":"https://doi.org/10.1109/ISPDS56360.2022.9874146","url":null,"abstract":"Due to the limitations imposed by and complexity of indoor environments, a low-cost and accurate indoor positioning system has not yet been designed. To address this issue, we constructed a fused indoor positioning algorithm based on the extended Kalman filter for WiFi and inertial measurement units (IMUs) using only a smartphone. To reduce the influence of WiFi signal fluctuation on fingerprint-based positioning, we used Gaussian process regression for denoising the data. We used our proposed improved clustering algorithm to reduce the matching amount in the positioning stage and increase the positioning accuracy. In terms of pedestrian dead reckoning (PDR) positioning, we designed a new and effective direction estimation algorithm integrating accelerometer and magnetometer, and we used an online step size estimation model to improve the accuracy of step size estimation. The experimental results showed that the average positioning error of the proposed fusion algorithm is 1.76 m, which was 55% lower than that using the WiFi network only, and 62% lower than using PDR only. Our findings showed that the fused positioning scheme based on WiFi and IMU can be used to effectively increase indoor positioning accuracy, and the proposed system is suitable for high-precision positioning scenarios.","PeriodicalId":280244,"journal":{"name":"2022 3rd International Conference on Information Science, Parallel and Distributed Systems (ISPDS)","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128899753","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 : 2022-07-22DOI: 10.1109/ISPDS56360.2022.9874164
Ning Zhou, Benyu Cui, Jianxin Zhou
Recently, the proton exchange membrane fuel cell (PEMFC) is of increasing interest to researchers and is considered to have a wide range of applications, because of its low pollution and high energy density. Remaining Useful Life (RUL) prediction is a major problem in driving the widespread use of PEMFC. This paper presents a transformer-based algorithm for RUL. The first step in this algorithm is to extract the periodicity and non-periodicity of the time series using time2vec. Then, the algorithm adds a convolutional network to the transformer to extract the temporal correlation and spatial correlation of the input time series. Moreover, we combine the handcrafted features with automatically learned features to boost the performance of the RUL prediction. The algorithm uses operational data from actual PEMFC vehicles for comparison experiments, and the prediction performance of our proposed algorithm outperforms prediction results of other algorithms.
{"title":"Transformer-based prediction of the RUL of PEMFC","authors":"Ning Zhou, Benyu Cui, Jianxin Zhou","doi":"10.1109/ISPDS56360.2022.9874164","DOIUrl":"https://doi.org/10.1109/ISPDS56360.2022.9874164","url":null,"abstract":"Recently, the proton exchange membrane fuel cell (PEMFC) is of increasing interest to researchers and is considered to have a wide range of applications, because of its low pollution and high energy density. Remaining Useful Life (RUL) prediction is a major problem in driving the widespread use of PEMFC. This paper presents a transformer-based algorithm for RUL. The first step in this algorithm is to extract the periodicity and non-periodicity of the time series using time2vec. Then, the algorithm adds a convolutional network to the transformer to extract the temporal correlation and spatial correlation of the input time series. Moreover, we combine the handcrafted features with automatically learned features to boost the performance of the RUL prediction. The algorithm uses operational data from actual PEMFC vehicles for comparison experiments, and the prediction performance of our proposed algorithm outperforms prediction results of other algorithms.","PeriodicalId":280244,"journal":{"name":"2022 3rd International Conference on Information Science, Parallel and Distributed Systems (ISPDS)","volume":"90 8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128936055","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：