Pub Date : 2021-12-10DOI: 10.1109/ICCSS53909.2021.9721967
Wanxing Xiao, Bo Yang
The development of automation and vehicle communication has enabled the cooperative control of vehicles at an intersection. In practice, in order to realize coordinated decisions and cooperative actions, information exchange among vehicles via a wireless network is susceptible to be affected by limited communication resources. In this paper, we consider a vehicle coordination and communication resource-aware problem under limited communication resources. Firstly, to address this issue, we propose a distributed model predictive control (DMPC) method with priority for each vehicle, which will reduce the impact of prediction consistency loss caused by communication delay. In addition, a prediction-based trigger mechanism is constructed for the proposed DMPC method, which predicts the usage of communication in advance and facilitates resource scheduling. Finally, we evaluated our scheme by the simulation of multi-vehicles, which demonstrates the effectiveness of communication saving while avoiding collisions and stop-deadlock.
{"title":"Cooperative Control of Intersection Connected Vehicles under Constrained Communication Resource","authors":"Wanxing Xiao, Bo Yang","doi":"10.1109/ICCSS53909.2021.9721967","DOIUrl":"https://doi.org/10.1109/ICCSS53909.2021.9721967","url":null,"abstract":"The development of automation and vehicle communication has enabled the cooperative control of vehicles at an intersection. In practice, in order to realize coordinated decisions and cooperative actions, information exchange among vehicles via a wireless network is susceptible to be affected by limited communication resources. In this paper, we consider a vehicle coordination and communication resource-aware problem under limited communication resources. Firstly, to address this issue, we propose a distributed model predictive control (DMPC) method with priority for each vehicle, which will reduce the impact of prediction consistency loss caused by communication delay. In addition, a prediction-based trigger mechanism is constructed for the proposed DMPC method, which predicts the usage of communication in advance and facilitates resource scheduling. Finally, we evaluated our scheme by the simulation of multi-vehicles, which demonstrates the effectiveness of communication saving while avoiding collisions and stop-deadlock.","PeriodicalId":435816,"journal":{"name":"2021 8th International Conference on Information, Cybernetics, and Computational Social Systems (ICCSS)","volume":"43 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":"122956511","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/ICCSS53909.2021.9722017
Kuan Lu, Zhijian Cheng, Hongru Ren, Renquan Lu
The concept and characteristics of the data space model of manufacturing enterprises in various countries are expounded, and a multi-dimensional data system architecture oriented to the data space of manufacturing enterprises is proposed. The effective analysis and processing of big data in manufacturing enterprises can provide them with more effective model building, integrated retrieval and intelligent management strategies, so as to reduce costs and increase efficiency. A systematic overview of the data space of the entire system and the entire value chain of the manufacturing enterprises is carried out. First, the three dimensions of the business domain, the processing domain and the modal domain are clarified; secondly, the methods of applying data processing at each stage in each domain are explained; finally, the advantages and importance of the data model are summarized.
{"title":"A Multidimensional System Architecture Oriented to the Data Space of Manufacturing Enterprises","authors":"Kuan Lu, Zhijian Cheng, Hongru Ren, Renquan Lu","doi":"10.1109/ICCSS53909.2021.9722017","DOIUrl":"https://doi.org/10.1109/ICCSS53909.2021.9722017","url":null,"abstract":"The concept and characteristics of the data space model of manufacturing enterprises in various countries are expounded, and a multi-dimensional data system architecture oriented to the data space of manufacturing enterprises is proposed. The effective analysis and processing of big data in manufacturing enterprises can provide them with more effective model building, integrated retrieval and intelligent management strategies, so as to reduce costs and increase efficiency. A systematic overview of the data space of the entire system and the entire value chain of the manufacturing enterprises is carried out. First, the three dimensions of the business domain, the processing domain and the modal domain are clarified; secondly, the methods of applying data processing at each stage in each domain are explained; finally, the advantages and importance of the data model are summarized.","PeriodicalId":435816,"journal":{"name":"2021 8th International Conference on Information, Cybernetics, and Computational Social Systems (ICCSS)","volume":"71 ","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114006505","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/ICCSS53909.2021.9721971
Fan Le, Hong Mo, Yinghui Meng
Lidar-based object detection is an important method of environment perception for autonomous driving. Due to the limitation of the inherent properties of lidar, the detection accuracy of obscured vehicles and distant objects is inferior, which causes the problem of missed detection. To address this problem, a lidar-millimeter wave radar information fusion multi-target detection method based on the unscented Kalman filter (UKF) and the covariance intersection (CI) algorithm was proposed in this article. Firstly, the UKF algorithm was applied to generate state estimations on the data collected by the sensor. Subsequently, the CI algorithm was introduced to form state fusion estimates. Finally, a simulation experiment platform was built based on MATLAB, and a comparison experiment with Joint Probabilistic Data Association (JPDA) and Gaussian mixture probability hypothesis density (GMPHD) algorithms were designed. The Generalized optimal sub-pattern assignment (GOSPA) indi-cators were adopted to evaluate the detection accuracy of each algorithm, and the effectiveness of the method was verified. The experimental results showed that UKF-CI had higher detection accuracy and provided accurate infor-mation for the decision-making part of the autonomous driving system, which guaranteed the stable operation of the autonomous driving system.
{"title":"Lidar-millimeter wave radar information fusion multi-target detection based on unscented Kalman filter and covariance intersection algorithm","authors":"Fan Le, Hong Mo, Yinghui Meng","doi":"10.1109/ICCSS53909.2021.9721971","DOIUrl":"https://doi.org/10.1109/ICCSS53909.2021.9721971","url":null,"abstract":"Lidar-based object detection is an important method of environment perception for autonomous driving. Due to the limitation of the inherent properties of lidar, the detection accuracy of obscured vehicles and distant objects is inferior, which causes the problem of missed detection. To address this problem, a lidar-millimeter wave radar information fusion multi-target detection method based on the unscented Kalman filter (UKF) and the covariance intersection (CI) algorithm was proposed in this article. Firstly, the UKF algorithm was applied to generate state estimations on the data collected by the sensor. Subsequently, the CI algorithm was introduced to form state fusion estimates. Finally, a simulation experiment platform was built based on MATLAB, and a comparison experiment with Joint Probabilistic Data Association (JPDA) and Gaussian mixture probability hypothesis density (GMPHD) algorithms were designed. The Generalized optimal sub-pattern assignment (GOSPA) indi-cators were adopted to evaluate the detection accuracy of each algorithm, and the effectiveness of the method was verified. The experimental results showed that UKF-CI had higher detection accuracy and provided accurate infor-mation for the decision-making part of the autonomous driving system, which guaranteed the stable operation of the autonomous driving system.","PeriodicalId":435816,"journal":{"name":"2021 8th International Conference on Information, Cybernetics, and Computational Social Systems (ICCSS)","volume":"49 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":"114553634","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/ICCSS53909.2021.9721949
Changqing Long, Guodong Zhang, Junhao Hu
This paper addresses the global asymptotic stabilization problem for a class of fuzzy inertial neural networks (FINNs) with infinite delays and handles with the FINNs directly by a non-reduced order strategy. By constructing Lyapunov functional and utilizing some analytical skills, new sufficient conditions are derived to assure the stabilization of the considered FINNs under the designed controller. Compared with the common neural networks, we introduce the fuzzy logics, inertial terms, time-varying coefficients and infinite delays into the considered model, which complements and improves on a number of existing publications. At last, two illustrative examples are given to demonstrate the validity of the theoretical outcomes.
{"title":"Stabilization of Fuzzy Inertial Neural Networks with Infinite Delays","authors":"Changqing Long, Guodong Zhang, Junhao Hu","doi":"10.1109/ICCSS53909.2021.9721949","DOIUrl":"https://doi.org/10.1109/ICCSS53909.2021.9721949","url":null,"abstract":"This paper addresses the global asymptotic stabilization problem for a class of fuzzy inertial neural networks (FINNs) with infinite delays and handles with the FINNs directly by a non-reduced order strategy. By constructing Lyapunov functional and utilizing some analytical skills, new sufficient conditions are derived to assure the stabilization of the considered FINNs under the designed controller. Compared with the common neural networks, we introduce the fuzzy logics, inertial terms, time-varying coefficients and infinite delays into the considered model, which complements and improves on a number of existing publications. At last, two illustrative examples are given to demonstrate the validity of the theoretical outcomes.","PeriodicalId":435816,"journal":{"name":"2021 8th International Conference on Information, Cybernetics, and Computational Social Systems (ICCSS)","volume":"50 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":"115451150","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}
Survival analysis (time-to-event analysis) is a set of statistic methods to analyze time-to-event data and is widely used in many fields such as economics, finance and medicine. One of the fundamental problems in survival analysis is to explore the relationship between the covariates and the survival time. Recently, with the development of deep learning-based techniques, various approaches have been proposed for survival analysis. To better handle the censoring, special cost functions or sophisticated network structures are usually designed for these methods. In this paper, a novel two-stage method is proposed to model the survival data. In the first stage, pseudo conditional probabilities are computed, which can act as the quantitative response variables in regression problems. In the second stage, with these pseudo values, a complicated survival analysis problem is transformed into a regression problem that can be effectively solved by broad learning system. The experimental results show that, with a flexible structure and a simple cost function, our proposed method has a better performance in handling the censored problems.
{"title":"BroadSurv: A Novel Broad Learning System-based Approach for Survival Analysis","authors":"Guangheng Wu, Junwei Duan, Jing Wang, Lu Wang, Cheng Dong, Changwei Lv","doi":"10.1109/ICCSS53909.2021.9721940","DOIUrl":"https://doi.org/10.1109/ICCSS53909.2021.9721940","url":null,"abstract":"Survival analysis (time-to-event analysis) is a set of statistic methods to analyze time-to-event data and is widely used in many fields such as economics, finance and medicine. One of the fundamental problems in survival analysis is to explore the relationship between the covariates and the survival time. Recently, with the development of deep learning-based techniques, various approaches have been proposed for survival analysis. To better handle the censoring, special cost functions or sophisticated network structures are usually designed for these methods. In this paper, a novel two-stage method is proposed to model the survival data. In the first stage, pseudo conditional probabilities are computed, which can act as the quantitative response variables in regression problems. In the second stage, with these pseudo values, a complicated survival analysis problem is transformed into a regression problem that can be effectively solved by broad learning system. The experimental results show that, with a flexible structure and a simple cost function, our proposed method has a better performance in handling the censored problems.","PeriodicalId":435816,"journal":{"name":"2021 8th International Conference on Information, Cybernetics, and Computational Social Systems (ICCSS)","volume":"6 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":"123364496","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/ICCSS53909.2021.9721989
Wenlin Li, Chuandong Li
Inspired by neural computing science, Spiking Neural Networks(SNNs), as the third generation of Artificial Neural Networks(ANNs), with its high biological interpretability, powerful time-space information processing ability and diverse spike coding method, has shown a great potential in pattern recognition, object detecting and data predicting. It has received extensive attention in the field of brain-inspired computing and machine learning. Utilizing spike trains as communication signals within the network is one of the advantages of spiking neural networks, which is the main way of information transmission between neurons in the brain. How to encode input information into spike signals for transmission in the network determines the working efficiency. In this paper, a spiking neural network based on the spike firing rate and temporal coding is proposed in the training and testing process respectively, and applied to the recognition of MNIST handwritten digital dataset, with an accuracy of 78.74%.
{"title":"An Image Recognizing method Based on Precise Moment of Spikes","authors":"Wenlin Li, Chuandong Li","doi":"10.1109/ICCSS53909.2021.9721989","DOIUrl":"https://doi.org/10.1109/ICCSS53909.2021.9721989","url":null,"abstract":"Inspired by neural computing science, Spiking Neural Networks(SNNs), as the third generation of Artificial Neural Networks(ANNs), with its high biological interpretability, powerful time-space information processing ability and diverse spike coding method, has shown a great potential in pattern recognition, object detecting and data predicting. It has received extensive attention in the field of brain-inspired computing and machine learning. Utilizing spike trains as communication signals within the network is one of the advantages of spiking neural networks, which is the main way of information transmission between neurons in the brain. How to encode input information into spike signals for transmission in the network determines the working efficiency. In this paper, a spiking neural network based on the spike firing rate and temporal coding is proposed in the training and testing process respectively, and applied to the recognition of MNIST handwritten digital dataset, with an accuracy of 78.74%.","PeriodicalId":435816,"journal":{"name":"2021 8th International Conference on Information, Cybernetics, and Computational Social Systems (ICCSS)","volume":"67 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":"129715312","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/ICCSS53909.2021.9721947
Yu Bai, Wenling Li, Bin Zhang
In this paper, we propose a distributed incremental quais-Newton (D-IQN) algorithm for multi-area power system state estimation (MASE). Maximum correntropy criterion (MCC) is used in objective function in order to address non-Gaussian noise. Incremental quais-Newton (IQN) is applied to solve state estimation in each area. In the inter-area communication networks, consensus+innovation strategy is adopted to form a distributed pattern. In this way, each area carries out a local state estimation with limited information exchange with its neighboring areas. As a fully distributed algorithm, no central coordinator is needed here. Based on this peer-to-peer communication paradigm, accurate estimation results are obtained and the privacy of each area remains well-preserved. Numerical experiments are carried out on 118-bus systems. The results show that the algorithm is effective for non-Gaussian noise and outperforms other methods such as distributed Broyden-Fletcher-Goldfarb-Shanno (BFGS), Gauss-Newton and WLS method.
{"title":"Distributed Incremental Quasi-Newton Algorithm for Power System State Estimation","authors":"Yu Bai, Wenling Li, Bin Zhang","doi":"10.1109/ICCSS53909.2021.9721947","DOIUrl":"https://doi.org/10.1109/ICCSS53909.2021.9721947","url":null,"abstract":"In this paper, we propose a distributed incremental quais-Newton (D-IQN) algorithm for multi-area power system state estimation (MASE). Maximum correntropy criterion (MCC) is used in objective function in order to address non-Gaussian noise. Incremental quais-Newton (IQN) is applied to solve state estimation in each area. In the inter-area communication networks, consensus+innovation strategy is adopted to form a distributed pattern. In this way, each area carries out a local state estimation with limited information exchange with its neighboring areas. As a fully distributed algorithm, no central coordinator is needed here. Based on this peer-to-peer communication paradigm, accurate estimation results are obtained and the privacy of each area remains well-preserved. Numerical experiments are carried out on 118-bus systems. The results show that the algorithm is effective for non-Gaussian noise and outperforms other methods such as distributed Broyden-Fletcher-Goldfarb-Shanno (BFGS), Gauss-Newton and WLS method.","PeriodicalId":435816,"journal":{"name":"2021 8th International Conference on Information, Cybernetics, and Computational Social Systems (ICCSS)","volume":"35 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":"124590271","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/ICCSS53909.2021.9722004
Xi Huang, Luheng Jia, Han Wang, Ke-bin Jia
Low-latency video applications are widely used in video communication, video surveillance and other real time scenarios, of which the low-latency video coding technique is the key component to reduce the coding complexity and transmission delay. The fixed-period intra refresh in video coding are capable of reducing inter-frame bit rate fluctuation and recover delivering error. In this work, we propose a novel fixed-period intra refresh method to further improve the coding efficiency and error resilience of encoded bitstream by rearranging the refreshing order according to the blocks importance-ranking joint considering reference importance using motion statistics and coding complexity leveraging rate-distortion cost of the encoding frame. Experimental results demonstrate that our proposed method obtains smoother bitrate and higher coding efficiency of up to 4.6% BD-rate reduction compared with previous method.
{"title":"Adaptive Intra Refresh For Low-Latency Video Coding","authors":"Xi Huang, Luheng Jia, Han Wang, Ke-bin Jia","doi":"10.1109/ICCSS53909.2021.9722004","DOIUrl":"https://doi.org/10.1109/ICCSS53909.2021.9722004","url":null,"abstract":"Low-latency video applications are widely used in video communication, video surveillance and other real time scenarios, of which the low-latency video coding technique is the key component to reduce the coding complexity and transmission delay. The fixed-period intra refresh in video coding are capable of reducing inter-frame bit rate fluctuation and recover delivering error. In this work, we propose a novel fixed-period intra refresh method to further improve the coding efficiency and error resilience of encoded bitstream by rearranging the refreshing order according to the blocks importance-ranking joint considering reference importance using motion statistics and coding complexity leveraging rate-distortion cost of the encoding frame. Experimental results demonstrate that our proposed method obtains smoother bitrate and higher coding efficiency of up to 4.6% BD-rate reduction compared with previous method.","PeriodicalId":435816,"journal":{"name":"2021 8th International Conference on Information, Cybernetics, and Computational Social Systems (ICCSS)","volume":"67 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":"124411055","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/ICCSS53909.2021.9721988
Zhijia Zhao, Jian Zhang, Jianing Zhang, Tao Zou
In this paper, we propose an adaptive neural network-based fault-tolerant control for the two-degree of freedom (DOF) helicopter system with actuator fault and output constraints. First, the radial basis function neural network is used to estimate the uncertainty of the system. Moreover, adaptive auxiliary parameters are used to compensate the actuator failure. And then, the barrier Lyapunov function is adopted to deal with the output constraints in the system. By analyzing the stability of Lyapunov function, it is strictly proved that the closed-loop system is semi-globally uniform and bounded, and under the combined action of actuator fault and output constraints, accurate tracking control performance is achieved. Finally, the simulation results in the 2-DOF helicopter system show the effectiveness of the control strategy.
{"title":"Adaptive Neural Network-Based Fault-Tolerant Control of 2-DOF Helicopter With Output Constraints","authors":"Zhijia Zhao, Jian Zhang, Jianing Zhang, Tao Zou","doi":"10.1109/ICCSS53909.2021.9721988","DOIUrl":"https://doi.org/10.1109/ICCSS53909.2021.9721988","url":null,"abstract":"In this paper, we propose an adaptive neural network-based fault-tolerant control for the two-degree of freedom (DOF) helicopter system with actuator fault and output constraints. First, the radial basis function neural network is used to estimate the uncertainty of the system. Moreover, adaptive auxiliary parameters are used to compensate the actuator failure. And then, the barrier Lyapunov function is adopted to deal with the output constraints in the system. By analyzing the stability of Lyapunov function, it is strictly proved that the closed-loop system is semi-globally uniform and bounded, and under the combined action of actuator fault and output constraints, accurate tracking control performance is achieved. Finally, the simulation results in the 2-DOF helicopter system show the effectiveness of the control strategy.","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":"127071485","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/ICCSS53909.2021.9721969
Mingwei Jia, Yun Dai, Danya Xu, Tao Yang, Yuan Yao, Yi Liu
In the (bio)chemical processes, traditional hardware sensors are difficult to directly measure the quality of critical products due to their time-varying, non-linear, and dynamic characteristics. This makes process soft sensor modeling methods important. Since the process variables can be regarded as natural graph data, this work introduces graphs in the soft sensor modeling area. A soft sensor model based on the graph neural network (GNN) is proposed. The model can learn the topological structure of graph data between each unit variable. Moreover, it characterizes variable relationships from the spatial and temporal dimensions to the output prediction by introducing the spatial-temporal convolutional layer. The effectiveness and advantages of the GNN-based soft sensor model are verified using a simulated fermentation process.
{"title":"Deep Graph Network for Process Soft Sensor Development","authors":"Mingwei Jia, Yun Dai, Danya Xu, Tao Yang, Yuan Yao, Yi Liu","doi":"10.1109/ICCSS53909.2021.9721969","DOIUrl":"https://doi.org/10.1109/ICCSS53909.2021.9721969","url":null,"abstract":"In the (bio)chemical processes, traditional hardware sensors are difficult to directly measure the quality of critical products due to their time-varying, non-linear, and dynamic characteristics. This makes process soft sensor modeling methods important. Since the process variables can be regarded as natural graph data, this work introduces graphs in the soft sensor modeling area. A soft sensor model based on the graph neural network (GNN) is proposed. The model can learn the topological structure of graph data between each unit variable. Moreover, it characterizes variable relationships from the spatial and temporal dimensions to the output prediction by introducing the spatial-temporal convolutional layer. The effectiveness and advantages of the GNN-based soft sensor model are verified using a simulated fermentation process.","PeriodicalId":435816,"journal":{"name":"2021 8th International Conference on Information, Cybernetics, and Computational Social Systems (ICCSS)","volume":"82 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":"133827351","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：