Pub Date : 2023-06-23DOI: 10.1109/ISAS59543.2023.10164523
Rendi Liu, Ju Jiang, Xiang Liu, Haowei Sun, Tingyu Ma
In this paper, a pitching control method based on Deep Deterministic Policy Gradient (DDPG) algorithm for carrier aircraft landing and descending stage is studied. DDPG controller takes pitch angle rate error, pitch angle error and altitude error as input, and output as elevator deflection, realizing the rapid pitch angle response of carrier-aircraft under different landing states. Compared with traditional PID controller, network training of Actor-Critic for DDPG attitude controller greatly improves the calculation efficiency of control quantity, and reduces the difficulty of parameter optimization. The simulation experiment in this paper was based on the F/A-18 aircraft aerodynamics model constructed in Matlab/Simulink, and the intensive learning and training environment built on PyCharm platform was used to realize the data interaction between the two platforms through UDP communication. The simulation results show that the attitude controller based on reinforcement learning designed in this paper has the characteristics of fast response speed and small dynamic error, and meets the control accuracy requirements in the experiment.
{"title":"Carrier Aircraft Landing Control Technology Based on Deep Reinforcement Learning","authors":"Rendi Liu, Ju Jiang, Xiang Liu, Haowei Sun, Tingyu Ma","doi":"10.1109/ISAS59543.2023.10164523","DOIUrl":"https://doi.org/10.1109/ISAS59543.2023.10164523","url":null,"abstract":"In this paper, a pitching control method based on Deep Deterministic Policy Gradient (DDPG) algorithm for carrier aircraft landing and descending stage is studied. DDPG controller takes pitch angle rate error, pitch angle error and altitude error as input, and output as elevator deflection, realizing the rapid pitch angle response of carrier-aircraft under different landing states. Compared with traditional PID controller, network training of Actor-Critic for DDPG attitude controller greatly improves the calculation efficiency of control quantity, and reduces the difficulty of parameter optimization. The simulation experiment in this paper was based on the F/A-18 aircraft aerodynamics model constructed in Matlab/Simulink, and the intensive learning and training environment built on PyCharm platform was used to realize the data interaction between the two platforms through UDP communication. The simulation results show that the attitude controller based on reinforcement learning designed in this paper has the characteristics of fast response speed and small dynamic error, and meets the control accuracy requirements in the experiment.","PeriodicalId":199115,"journal":{"name":"2023 6th International Symposium on Autonomous Systems (ISAS)","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-06-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124661906","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 : 2023-06-23DOI: 10.1109/ISAS59543.2023.10164609
Maomao Zhao, Shaojie Zhang, Bin Jiang
A novel potential function multi-agent deep deterministic policy gradient (PF-MADDPG) algorithm is proposed for the multi-agent Attacker-Defender-Target (ADT). A multi-agent continuous state space and a continuous action space are established. The potential function rewards of target and defenders are designed to accelerate the game confrontation training speed, and the MADDPG algorithm is utilized to obtain effective strategies, so as to describe the influence of different actions on attackers. Finally, simulations are given to verify the effectiveness of the proposed PF-MADDPG algorithm.
{"title":"Multi-Agent Cooperative Attacker-Defender-Target Task Decision Based on PF-MADDPG","authors":"Maomao Zhao, Shaojie Zhang, Bin Jiang","doi":"10.1109/ISAS59543.2023.10164609","DOIUrl":"https://doi.org/10.1109/ISAS59543.2023.10164609","url":null,"abstract":"A novel potential function multi-agent deep deterministic policy gradient (PF-MADDPG) algorithm is proposed for the multi-agent Attacker-Defender-Target (ADT). A multi-agent continuous state space and a continuous action space are established. The potential function rewards of target and defenders are designed to accelerate the game confrontation training speed, and the MADDPG algorithm is utilized to obtain effective strategies, so as to describe the influence of different actions on attackers. Finally, simulations are given to verify the effectiveness of the proposed PF-MADDPG algorithm.","PeriodicalId":199115,"journal":{"name":"2023 6th International Symposium on Autonomous Systems (ISAS)","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-06-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133414032","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 : 2023-06-23DOI: 10.1109/ISAS59543.2023.10164296
Hengjie Dai, Jianhua Lyu, Mejed Jebali
In the process of industrial production and to ensure the production order, it is necessary to monitor the process in real time, detect errors and take action in advance to reduce losses. Failure Mode and Effects Analysis (FMEA) is a systematic activity to analyze product modules, parts and various operations in the production process to identify potential failure modes and analyze their possible consequences. This leads to necessary actions being taken in advance to improve product quality and reliability. Efficient management of FMEA data is beneficial for controlling the production process and improving production quality. Based on the failure mode and effects analysis (FMEA) data of industrial systems, this paper builds a knowledge graph of failure modes and designs, and develops the corresponding modules for the management functions, including knowledge graph creation, knowledge graph storage, and knowledge graph retrieval. First, the ontology structure of the failure mode is designed in terms of the failure mode of industrial systems. Second, the facts are extracted from the unstructured data in FMEA, the structured data is cleaned, the abnormal data is eliminated, and the missing data is recovered. Third, according to the correlation between the pattern level ontology, the knowledge graph triplet is created and the FMEA knowledge graph is constructed; then the storage function of the FMEA knowledge graph is designed and implemented based on the graph database neo4j; finally, the KNN algorithm for the similarity search in the FMEA knowledge graph is proposed.
在工业生产过程中，为了保证生产秩序，需要对过程进行实时监控，发现错误并提前采取行动，以减少损失。失效模式与影响分析(Failure Mode and Effects Analysis, FMEA)是对产品模块、零部件和生产过程中的各种操作进行分析，以识别潜在失效模式并分析其可能后果的系统活动。这导致提前采取必要的措施来提高产品质量和可靠性。对FMEA数据进行有效的管理，有利于控制生产过程，提高生产质量。基于工业系统失效模式与影响分析(FMEA)数据，构建了失效模式知识图谱并进行了设计，开发了相应的管理功能模块，包括知识图谱创建、知识图谱存储和知识图谱检索。首先，从工业系统的失效模式出发，设计了失效模式本体结构;其次，对FMEA中的非结构化数据进行事实提取，对结构化数据进行清洗，剔除异常数据，恢复缺失数据;第三，根据模式级本体之间的关联关系，创建知识图谱三元组，构建FMEA知识图谱;然后基于图形数据库neo4j设计并实现了FMEA知识图谱的存储功能;最后，提出了用于FMEA知识图相似性搜索的KNN算法。
{"title":"Construction and management of industrial system failure mode knowledge graph","authors":"Hengjie Dai, Jianhua Lyu, Mejed Jebali","doi":"10.1109/ISAS59543.2023.10164296","DOIUrl":"https://doi.org/10.1109/ISAS59543.2023.10164296","url":null,"abstract":"In the process of industrial production and to ensure the production order, it is necessary to monitor the process in real time, detect errors and take action in advance to reduce losses. Failure Mode and Effects Analysis (FMEA) is a systematic activity to analyze product modules, parts and various operations in the production process to identify potential failure modes and analyze their possible consequences. This leads to necessary actions being taken in advance to improve product quality and reliability. Efficient management of FMEA data is beneficial for controlling the production process and improving production quality. Based on the failure mode and effects analysis (FMEA) data of industrial systems, this paper builds a knowledge graph of failure modes and designs, and develops the corresponding modules for the management functions, including knowledge graph creation, knowledge graph storage, and knowledge graph retrieval. First, the ontology structure of the failure mode is designed in terms of the failure mode of industrial systems. Second, the facts are extracted from the unstructured data in FMEA, the structured data is cleaned, the abnormal data is eliminated, and the missing data is recovered. Third, according to the correlation between the pattern level ontology, the knowledge graph triplet is created and the FMEA knowledge graph is constructed; then the storage function of the FMEA knowledge graph is designed and implemented based on the graph database neo4j; finally, the KNN algorithm for the similarity search in the FMEA knowledge graph is proposed.","PeriodicalId":199115,"journal":{"name":"2023 6th International Symposium on Autonomous Systems (ISAS)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-06-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133015477","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 : 2023-06-23DOI: 10.1109/ISAS59543.2023.10164567
Mingzheng Yu, Ticao Jiao, Housheng Zhang, Lei Wang
The aim of this paper is to investigate the issue of stochastically exponential stability in positive nonlinear impulsive switching systems that involve random impulses. The random impulses means that the randomness of impulsive densities and the occurrence of the impulses is a Markov chain. The occurrence instants of impulses also satisfy average impulsive intervals. This research employs the multiple max-separable Lyapunov function method to investigate two cases: asynchronous/synchronous switching and impulses. Then the stochastic exponential stability conditions are given. Finally, to demonstrate the validity of the theoretical results, two illustrative examples are provided.
{"title":"Stability Analysis for Nonlinear Switched Positive Systems With Markov Chain Impulses","authors":"Mingzheng Yu, Ticao Jiao, Housheng Zhang, Lei Wang","doi":"10.1109/ISAS59543.2023.10164567","DOIUrl":"https://doi.org/10.1109/ISAS59543.2023.10164567","url":null,"abstract":"The aim of this paper is to investigate the issue of stochastically exponential stability in positive nonlinear impulsive switching systems that involve random impulses. The random impulses means that the randomness of impulsive densities and the occurrence of the impulses is a Markov chain. The occurrence instants of impulses also satisfy average impulsive intervals. This research employs the multiple max-separable Lyapunov function method to investigate two cases: asynchronous/synchronous switching and impulses. Then the stochastic exponential stability conditions are given. Finally, to demonstrate the validity of the theoretical results, two illustrative examples are provided.","PeriodicalId":199115,"journal":{"name":"2023 6th International Symposium on Autonomous Systems (ISAS)","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-06-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116595687","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 : 2023-06-23DOI: 10.1109/ISAS59543.2023.10164329
Chaoran Wang, Hai Zhu, Xiaozhou Zhu, C. Wu, Wen Yao, Xiaoqian Chen
This paper proposes a task scheduling framework for autonomous navigation of micro aerial vehicles (MAVs) in unknown environments. Currently, the dominant approach for task scheduling in MAV systems typically relies on the finite state machine (FSM), which presents limitations in scaling the system functionality due to the coupled relationship between modules. We propose a generic task scheduling framework for MAVs based on behavior trees (BTs) to address this challenge. A blackboard is built as a state data management center, decoupling different modules of the MAV so that the various functional modules can operate independently. In addition, we set up standardized interfaces for the modules, thus the behavior tree can quickly connect the modules. This framework enables MAVs to perform autonomous flight scheduling tasks and supports the rapid expansion of system functions. Finally, we validated the effectiveness of the framework with real-world experiments on a customized MAV.
{"title":"Task Management for Autonomous Flights of Micro Aerial Vehicles: A Behavior Tree Approach","authors":"Chaoran Wang, Hai Zhu, Xiaozhou Zhu, C. Wu, Wen Yao, Xiaoqian Chen","doi":"10.1109/ISAS59543.2023.10164329","DOIUrl":"https://doi.org/10.1109/ISAS59543.2023.10164329","url":null,"abstract":"This paper proposes a task scheduling framework for autonomous navigation of micro aerial vehicles (MAVs) in unknown environments. Currently, the dominant approach for task scheduling in MAV systems typically relies on the finite state machine (FSM), which presents limitations in scaling the system functionality due to the coupled relationship between modules. We propose a generic task scheduling framework for MAVs based on behavior trees (BTs) to address this challenge. A blackboard is built as a state data management center, decoupling different modules of the MAV so that the various functional modules can operate independently. In addition, we set up standardized interfaces for the modules, thus the behavior tree can quickly connect the modules. This framework enables MAVs to perform autonomous flight scheduling tasks and supports the rapid expansion of system functions. Finally, we validated the effectiveness of the framework with real-world experiments on a customized MAV.","PeriodicalId":199115,"journal":{"name":"2023 6th International Symposium on Autonomous Systems (ISAS)","volume":"273 ","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-06-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120887491","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 : 2023-06-23DOI: 10.1109/ISAS59543.2023.10164356
Long Zhao, Rongjie Liu, Shi-Yu Li, Xiangyu Wang, De Bao
This paper proposes a strategy for improving the correct diagnosis of epilepsy based on electroencephalogram (EEG) using a spatio-temporal variable structure graph convolutional neural network. Specifically, this method is called the variable-structure graph convolutional neural network (VGCRN), which is derived by combining spatial information and noise removal through variable structured graph convolutional neural network and temporal information through recurrent neural network. Despite the potential benefits of EEG for diagnosing and monitoring neurological conditions, the low signal-to-noise ratio often hinders timely and accurate diagnosis in many clinical cases. Previous research on EEG data classification has mainly focused on extracting features from the time or frequency domain, disregarding the spatial features among electrodes. EEG can be viewed as a structured time series, consisting of multivariate time series data with prior information provided by the spatial location of electrodes on the patient’s scalp. Spatial information is just as crucial as time or frequency-domain information, but introducing unconstrained spatial features in topological map structures can result in noise and the aggregation of irrelevant information by nodes. The proposed method in this paper can better leverage the spatial and intrinsic temporal information of brain waves while reducing noise, thus enhancing the robustness and accuracy of the model.
{"title":"Spatio-Temporal Variable Structure Graph Neural Network for EEG Data Classification","authors":"Long Zhao, Rongjie Liu, Shi-Yu Li, Xiangyu Wang, De Bao","doi":"10.1109/ISAS59543.2023.10164356","DOIUrl":"https://doi.org/10.1109/ISAS59543.2023.10164356","url":null,"abstract":"This paper proposes a strategy for improving the correct diagnosis of epilepsy based on electroencephalogram (EEG) using a spatio-temporal variable structure graph convolutional neural network. Specifically, this method is called the variable-structure graph convolutional neural network (VGCRN), which is derived by combining spatial information and noise removal through variable structured graph convolutional neural network and temporal information through recurrent neural network. Despite the potential benefits of EEG for diagnosing and monitoring neurological conditions, the low signal-to-noise ratio often hinders timely and accurate diagnosis in many clinical cases. Previous research on EEG data classification has mainly focused on extracting features from the time or frequency domain, disregarding the spatial features among electrodes. EEG can be viewed as a structured time series, consisting of multivariate time series data with prior information provided by the spatial location of electrodes on the patient’s scalp. Spatial information is just as crucial as time or frequency-domain information, but introducing unconstrained spatial features in topological map structures can result in noise and the aggregation of irrelevant information by nodes. The proposed method in this paper can better leverage the spatial and intrinsic temporal information of brain waves while reducing noise, thus enhancing the robustness and accuracy of the model.","PeriodicalId":199115,"journal":{"name":"2023 6th International Symposium on Autonomous Systems (ISAS)","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-06-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130365610","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 : 2023-06-23DOI: 10.1109/ISAS59543.2023.10164390
Yuran Li, X. Zhai, Ran Wang, Jing Zhu, Chenghua Yao
The combination of cognitive radio networks and unmanned aerial vehicles (UAVs) can overcome the scarcity of spectrum for wireless networks and provide important benefits for large-scale deployment of UAVs. However, energy constraint is an inevitable challenge for UAV communication. In order to prolong the lifetime of UAVs, we apply the accumulating priority queue (APQ) discipline for the first time to the queuing charging process of UAVs in cognitive radio networks. The APQ is based on the classic priority queue with the accumulation rate added to it, which means that the priority of a user is a function of the waiting time. By analyzing the values for different arrival rates and priorities, we discuss the factors that affect the average waiting times and key performance indicators. We also consider the related optimization problem and propose an improved bisection algorithm. According to the requirements of different classes of users, we can adjust the accumulation rates and arrival rates to meet the performance indicators.
{"title":"Accumulating Priority Queue for Charging of Unmanned Aerial Vehicles in Cognitive Radio Networks","authors":"Yuran Li, X. Zhai, Ran Wang, Jing Zhu, Chenghua Yao","doi":"10.1109/ISAS59543.2023.10164390","DOIUrl":"https://doi.org/10.1109/ISAS59543.2023.10164390","url":null,"abstract":"The combination of cognitive radio networks and unmanned aerial vehicles (UAVs) can overcome the scarcity of spectrum for wireless networks and provide important benefits for large-scale deployment of UAVs. However, energy constraint is an inevitable challenge for UAV communication. In order to prolong the lifetime of UAVs, we apply the accumulating priority queue (APQ) discipline for the first time to the queuing charging process of UAVs in cognitive radio networks. The APQ is based on the classic priority queue with the accumulation rate added to it, which means that the priority of a user is a function of the waiting time. By analyzing the values for different arrival rates and priorities, we discuss the factors that affect the average waiting times and key performance indicators. We also consider the related optimization problem and propose an improved bisection algorithm. According to the requirements of different classes of users, we can adjust the accumulation rates and arrival rates to meet the performance indicators.","PeriodicalId":199115,"journal":{"name":"2023 6th International Symposium on Autonomous Systems (ISAS)","volume":"22 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-06-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129391541","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}
In recent years, indoor inertial navigation technology based on pedestrian dead reckoning (PDR) has been widely promoted. Traditional methods often use auxiliary facilities or environmental constraints to suppress PDR heading cumulative errors, but these auxiliary means restrict the application scope of PDR. PDR based on deep learning fills the need for external information dependence, but the heading estimation accuracy is low and the adaptability is poor. To address this problem, an optimized adaptive multitask loss layer based on uncertain weighting is proposed, which constrains the weight of position and attitude estimation in the overall prediction task and dynamically adjusts it adaptively in different stages to enhance attitude estimation capability. A PDR algorithm based on an end-to-end joint residual neural network and bidirectional long short-term memory network is designed to improve the algorithm’s generalization ability. The original inertial navigation data is processed by segmentation and coordinate normalization and is used as input to the deep learning model to detect features and predict trajectories, achieving accurate indoor pedestrian inertial navigation. Finally, the navigation performance of the proposed algorithm is validated in experiments of walking, running, and mixed gait patterns. The results show that the positioning accuracy of the proposed algorithm is better than that of traditional PDR methods and the RONIN algorithm based on deep learning. The positioning errors in walking, running, and mixed gait patterns are reduced by 21.07%, 10.34%, and 32.15%, respectively, compared to the RONIN algorithm.
{"title":"Deep Learning Pedestrian Navigation Method Based on Multi-task Loss Function*","authors":"Tao Wang, Jizhou Lai, Cheng Yuan, Jingyi Zhu, Qianqian Zhu, Pin Lyu","doi":"10.1109/ISAS59543.2023.10164517","DOIUrl":"https://doi.org/10.1109/ISAS59543.2023.10164517","url":null,"abstract":"In recent years, indoor inertial navigation technology based on pedestrian dead reckoning (PDR) has been widely promoted. Traditional methods often use auxiliary facilities or environmental constraints to suppress PDR heading cumulative errors, but these auxiliary means restrict the application scope of PDR. PDR based on deep learning fills the need for external information dependence, but the heading estimation accuracy is low and the adaptability is poor. To address this problem, an optimized adaptive multitask loss layer based on uncertain weighting is proposed, which constrains the weight of position and attitude estimation in the overall prediction task and dynamically adjusts it adaptively in different stages to enhance attitude estimation capability. A PDR algorithm based on an end-to-end joint residual neural network and bidirectional long short-term memory network is designed to improve the algorithm’s generalization ability. The original inertial navigation data is processed by segmentation and coordinate normalization and is used as input to the deep learning model to detect features and predict trajectories, achieving accurate indoor pedestrian inertial navigation. Finally, the navigation performance of the proposed algorithm is validated in experiments of walking, running, and mixed gait patterns. The results show that the positioning accuracy of the proposed algorithm is better than that of traditional PDR methods and the RONIN algorithm based on deep learning. The positioning errors in walking, running, and mixed gait patterns are reduced by 21.07%, 10.34%, and 32.15%, respectively, compared to the RONIN algorithm.","PeriodicalId":199115,"journal":{"name":"2023 6th International Symposium on Autonomous Systems (ISAS)","volume":"46 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-06-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126431932","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 : 2023-06-23DOI: 10.1109/ISAS59543.2023.10164606
Rong Zhao, Lu Liu
This paper investigates the problem of $mathcal{H}{_}{-} / mathcal{H}{_}{infty}$ fault detection (FD) filter design for continuous-time polytopic uncertain linear systems in the finite-frequency (FF) domain. By assuming that both disturbances and faults are restricted to FF ranges, we are interested in designing an FD filter such that the resulting filtering error system (FES) is both sensitive to faults and robust against disturbances. By using the generalized Kalman-Yakubovič-Popov (KYP) lemma, Projection lemma, and some elegant convexification procedures, sufficient conditions for synthesis of the FD filter are established by solving an optimization problem in the form of linear matrix inequalities (LMIs). Finally, simulation studies are provided to validate the effectiveness of the proposed filtering approach.
{"title":"Fault Detection Filter Design for Polytopic Uncertain Systems in Finite-Frequency Domain","authors":"Rong Zhao, Lu Liu","doi":"10.1109/ISAS59543.2023.10164606","DOIUrl":"https://doi.org/10.1109/ISAS59543.2023.10164606","url":null,"abstract":"This paper investigates the problem of $mathcal{H}{_}{-} / mathcal{H}{_}{infty}$ fault detection (FD) filter design for continuous-time polytopic uncertain linear systems in the finite-frequency (FF) domain. By assuming that both disturbances and faults are restricted to FF ranges, we are interested in designing an FD filter such that the resulting filtering error system (FES) is both sensitive to faults and robust against disturbances. By using the generalized Kalman-Yakubovič-Popov (KYP) lemma, Projection lemma, and some elegant convexification procedures, sufficient conditions for synthesis of the FD filter are established by solving an optimization problem in the form of linear matrix inequalities (LMIs). Finally, simulation studies are provided to validate the effectiveness of the proposed filtering approach.","PeriodicalId":199115,"journal":{"name":"2023 6th International Symposium on Autonomous Systems (ISAS)","volume":"71 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-06-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127321121","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 : 2023-06-23DOI: 10.1109/ISAS59543.2023.10164341
Chunlei Shi, D. Niu, Hao Gong, Mei Zhang, Zhan Cao, Yulong Jin
This paper proposes a lightweight person re-identification network that incorporates a progressive attention mechanism The network aims to address the low accuracy issue in person re-identification caused by various factors such as different viewing angles, poses, illumination conditions, occlusions, and low image resolutions. Additionally, the network design takes into consideration the need for lightweight model deployment in practical scenarios. To improve the network’s performance using limited training data, data augmentation techniques are employed to expand the training dataset and enhance the robustness of the network model. The Resnet-50 architecture serves as the backbone network, and a feature shunt structure with depthwise separable convolutions is introduced to reduce computational parameters and accelerate person retrieval inference speed. Furthermore, the feature extraction and embedding processes are separated, and a progressive attention module is introduced. This module gradually segments the features into local blocks of different granularity, allowing for the learning of discriminative features at each granularity level. This progressive approach enhances the network’s ability to perceive foreground information from coarse to fine levels and improves feature matching capability. To supervise the model, a triplet loss function is utilized, specifically designed to address challenging samples. This loss function helps reduce intra-class variations while increasing inter-class separability. The efficacy of the proposed method in person re-identification is substantiated by conducting experimental evaluation on both the Market-1501 and DukeMTMC-ReID datasets. The experimental results demonstrate that the method achieves mAP indices of 88.1% and 79.1% on the respective datasets, providing strong evidence for its effectiveness in addressing the challenges of person re-identification.
{"title":"Person Re-identification Lightweight Network Based on Progressive Attention Mechanism","authors":"Chunlei Shi, D. Niu, Hao Gong, Mei Zhang, Zhan Cao, Yulong Jin","doi":"10.1109/ISAS59543.2023.10164341","DOIUrl":"https://doi.org/10.1109/ISAS59543.2023.10164341","url":null,"abstract":"This paper proposes a lightweight person re-identification network that incorporates a progressive attention mechanism The network aims to address the low accuracy issue in person re-identification caused by various factors such as different viewing angles, poses, illumination conditions, occlusions, and low image resolutions. Additionally, the network design takes into consideration the need for lightweight model deployment in practical scenarios. To improve the network’s performance using limited training data, data augmentation techniques are employed to expand the training dataset and enhance the robustness of the network model. The Resnet-50 architecture serves as the backbone network, and a feature shunt structure with depthwise separable convolutions is introduced to reduce computational parameters and accelerate person retrieval inference speed. Furthermore, the feature extraction and embedding processes are separated, and a progressive attention module is introduced. This module gradually segments the features into local blocks of different granularity, allowing for the learning of discriminative features at each granularity level. This progressive approach enhances the network’s ability to perceive foreground information from coarse to fine levels and improves feature matching capability. To supervise the model, a triplet loss function is utilized, specifically designed to address challenging samples. This loss function helps reduce intra-class variations while increasing inter-class separability. The efficacy of the proposed method in person re-identification is substantiated by conducting experimental evaluation on both the Market-1501 and DukeMTMC-ReID datasets. The experimental results demonstrate that the method achieves mAP indices of 88.1% and 79.1% on the respective datasets, providing strong evidence for its effectiveness in addressing the challenges of person re-identification.","PeriodicalId":199115,"journal":{"name":"2023 6th International Symposium on Autonomous Systems (ISAS)","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-06-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126946214","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：