Dear Editor, This letter proposes a symmetry-preserving dual-stream graph neural network (SDGNN) for precise representation learning to an undirected weighted graph (UWG). Although existing graph neural networks (GNNs) are influential instruments for representation learning to a UWG, they invariably adopt a unique node feature matrix for illustrating the sole node set of a UWG. Such a modeling strategy can limit the representation learning ability due to the diminished feature space. To this end, the proposed SDGNN innovatively adopts the following two-fold ideas: 1) Building a dual-stream graph learning framework that tolerates multiple node feature matrices for boosting the representation learning ability; 2) Integrating a symmetry regularization term into the learning objective for implying the equality constraint among its multiple node feature matrices, which exemplifies a graph's intrinsic symmetry and prompts learning the multiple node embeddings jointly. Experiments on six real-world UWG datasets indicate that the proposed SDGNN has superior performance in addressing the task of missing link estimation compared with the state-of-the-art baselines.
{"title":"SDGNN: Symmetry-Preserving Dual-Stream Graph Neural Networks","authors":"Jiufang Chen;Ye Yuan;Xin Luo","doi":"10.1109/JAS.2024.124410","DOIUrl":"https://doi.org/10.1109/JAS.2024.124410","url":null,"abstract":"Dear Editor, This letter proposes a symmetry-preserving dual-stream graph neural network (SDGNN) for precise representation learning to an undirected weighted graph (UWG). Although existing graph neural networks (GNNs) are influential instruments for representation learning to a UWG, they invariably adopt a unique node feature matrix for illustrating the sole node set of a UWG. Such a modeling strategy can limit the representation learning ability due to the diminished feature space. To this end, the proposed SDGNN innovatively adopts the following two-fold ideas: 1) Building a dual-stream graph learning framework that tolerates multiple node feature matrices for boosting the representation learning ability; 2) Integrating a symmetry regularization term into the learning objective for implying the equality constraint among its multiple node feature matrices, which exemplifies a graph's intrinsic symmetry and prompts learning the multiple node embeddings jointly. Experiments on six real-world UWG datasets indicate that the proposed SDGNN has superior performance in addressing the task of missing link estimation compared with the state-of-the-art baselines.","PeriodicalId":54230,"journal":{"name":"Ieee-Caa Journal of Automatica Sinica","volume":null,"pages":null},"PeriodicalIF":11.8,"publicationDate":"2024-06-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10555245","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141315265","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Dear Editor, Tracking control in networked environment is a very challenging problem due to the contradiction of rapid response to the time-varying signal and the inevitable delay introduced by networks. This letter has proposed several fuzzy-inverse-model-based network tracking control frameworks which are helpful in handling the system with nonlinear dynamics and uncertainties. The control frameworks have adopted different strategies such as feedback correction, internal model structure and adaptive technology. Simulations have proved the validity of the strategies. Moreover, the combination of two or more technologies can greatly improve the control performance.
{"title":"Fuzzy-Inverse-Model-Based Networked Tracking Control Frameworks of Time-Varying Signals","authors":"Shiwen Tong;Dianwei Qian;Keya Yuan;Dexin Liu;Yuan Li;Jiancheng Zhang","doi":"10.1109/JAS.2024.124293","DOIUrl":"https://doi.org/10.1109/JAS.2024.124293","url":null,"abstract":"Dear Editor, Tracking control in networked environment is a very challenging problem due to the contradiction of rapid response to the time-varying signal and the inevitable delay introduced by networks. This letter has proposed several fuzzy-inverse-model-based network tracking control frameworks which are helpful in handling the system with nonlinear dynamics and uncertainties. The control frameworks have adopted different strategies such as feedback correction, internal model structure and adaptive technology. Simulations have proved the validity of the strategies. Moreover, the combination of two or more technologies can greatly improve the control performance.","PeriodicalId":54230,"journal":{"name":"Ieee-Caa Journal of Automatica Sinica","volume":null,"pages":null},"PeriodicalIF":11.8,"publicationDate":"2024-06-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10555240","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141315266","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Dear Editor, In this letter, an output tracking control problem of uncertain cyber-physical systems (CPSs) is considered in the perspective of high-order fully actuated (HOFA) system theory, where a lumped disturbance is used to denote the total uncertainties containing parameters perturbations and external disturbances. A disturbance observer-based HOFA predictive control (DOB-HOFAPC) is adopted to achieve the desired tracking control performance and compensate for the communication delays in the forward and backward channels. The further discussion gives a criterion to analyze the tracking performance and stability of closed-loop CPSs. An example of long distance power transmission line is shown to verify the feasibility of the proposed DOB-HOFAPC.
{"title":"Disturbance Observer-Based Predictive Tracking Control of Uncertain HOFA Cyber-Physical Systems","authors":"Da-Wei Zhang;Guo-Ping Liu","doi":"10.1109/JAS.2023.124080","DOIUrl":"https://doi.org/10.1109/JAS.2023.124080","url":null,"abstract":"Dear Editor, In this letter, an output tracking control problem of uncertain cyber-physical systems (CPSs) is considered in the perspective of high-order fully actuated (HOFA) system theory, where a lumped disturbance is used to denote the total uncertainties containing parameters perturbations and external disturbances. A disturbance observer-based HOFA predictive control (DOB-HOFAPC) is adopted to achieve the desired tracking control performance and compensate for the communication delays in the forward and backward channels. The further discussion gives a criterion to analyze the tracking performance and stability of closed-loop CPSs. An example of long distance power transmission line is shown to verify the feasibility of the proposed DOB-HOFAPC.","PeriodicalId":54230,"journal":{"name":"Ieee-Caa Journal of Automatica Sinica","volume":null,"pages":null},"PeriodicalIF":11.8,"publicationDate":"2024-06-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10555244","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141315150","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Yuhan Zhang;Zidong Wang;Lei Zou;Yun Chen;Guoping Lu
This paper concerns ultimately bounded output-feedback control problems for networked systems with unknown nonlinear dynamics. Sensor-to-observer signal transmission is facilitated over networks that has communication constraints. These transmissions are carried out over an unreliable communication channel. In order to enhance the utilization rate of measurement data, a buffer-aided strategy is novelly employed to store historical measurements when communication networks are inaccessible. Using the neural network technique, a novel observer-based controller is introduced to address effects of signal transmission behaviors and unknown nonlinear dynamics. Through the application of stochastic analysis and Lyapunov stability, a joint framework is constructed for analyzing resultant system performance under the introduced controller. Subsequently, existence conditions for the desired output-feedback controller are delineated. The required parameters for the observer-based controller are then determined by resolving some specific matrix inequalities. Finally, a simulation example is showcased to confirm method efficacy.
{"title":"Ultimately Bounded Output Feedback Control for Networked Nonlinear Systems With Unreliable Communication Channel: A Buffer-Aided Strategy","authors":"Yuhan Zhang;Zidong Wang;Lei Zou;Yun Chen;Guoping Lu","doi":"10.1109/JAS.2024.124314","DOIUrl":"https://doi.org/10.1109/JAS.2024.124314","url":null,"abstract":"This paper concerns ultimately bounded output-feedback control problems for networked systems with unknown nonlinear dynamics. Sensor-to-observer signal transmission is facilitated over networks that has communication constraints. These transmissions are carried out over an unreliable communication channel. In order to enhance the utilization rate of measurement data, a buffer-aided strategy is novelly employed to store historical measurements when communication networks are inaccessible. Using the neural network technique, a novel observer-based controller is introduced to address effects of signal transmission behaviors and unknown nonlinear dynamics. Through the application of stochastic analysis and Lyapunov stability, a joint framework is constructed for analyzing resultant system performance under the introduced controller. Subsequently, existence conditions for the desired output-feedback controller are delineated. The required parameters for the observer-based controller are then determined by resolving some specific matrix inequalities. Finally, a simulation example is showcased to confirm method efficacy.","PeriodicalId":54230,"journal":{"name":"Ieee-Caa Journal of Automatica Sinica","volume":null,"pages":null},"PeriodicalIF":11.8,"publicationDate":"2024-06-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141315252","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Yingjiang Guo;Wenying Xu;Haodong Wang;Jianquan Lu;Shengli Du
Dear Editor, Over the past decades, cooperative control and distributed optimization have gained significant research attention due to their broad applications such as signal processing, robotics, and social networks [1], [2]. As a fundamental component of distributed control and optimization, the issue of average consensus has become a recurring topic of interest [3], [4]. To achieve average consensus, it is essential to establish a distributed algorithm with local information under which each node is able to adjust its own behavior by exchanging information with its neighbors instead of relying on a central node.
{"title":"Privacy-Preserving Average Consensus Algorithm Under Round-Robin Scheduling Protocol","authors":"Yingjiang Guo;Wenying Xu;Haodong Wang;Jianquan Lu;Shengli Du","doi":"10.1109/JAS.2023.123921","DOIUrl":"https://doi.org/10.1109/JAS.2023.123921","url":null,"abstract":"Dear Editor, Over the past decades, cooperative control and distributed optimization have gained significant research attention due to their broad applications such as signal processing, robotics, and social networks [1], [2]. As a fundamental component of distributed control and optimization, the issue of average consensus has become a recurring topic of interest [3], [4]. To achieve average consensus, it is essential to establish a distributed algorithm with local information under which each node is able to adjust its own behavior by exchanging information with its neighbors instead of relying on a central node.","PeriodicalId":54230,"journal":{"name":"Ieee-Caa Journal of Automatica Sinica","volume":null,"pages":null},"PeriodicalIF":11.8,"publicationDate":"2024-06-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10555246","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141315151","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Yufeng Lian;Xingtian Xiao;Jiliang Zhang;Long Jin;Junzhi Yu;Zhongbo Sun
This paper presents a distributed scheme with limited communications, aiming to achieve cooperative motion control for multiple omnidirectional mobile manipulators (MOMMs). The proposed scheme extends the existing single-agent motion control to cater to scenarios involving the cooperative operation of MOMMs. Specifically, squeeze-free cooperative load transportation is achieved for the end-effectors of MOMMs by incorporating cooperative repetitive motion planning (CRMP), while guiding each individual to desired poses. Then, the distributed scheme is formulated as a time-varying quadratic programming (QP) and solved online utilizing a noise-tolerant zeroing neural network (NTZNN). Theoretical analysis shows that the NTZNN model converges globally to the optimal solution of QP in the presence of noise. Finally, the effectiveness of the control design is demon-strated by numerical simulations and physical platform experiments.
{"title":"Neural Dynamics for Cooperative Motion Control of Omnidirectional Mobile Manipulators in the Presence of Noises: A Distributed Approach","authors":"Yufeng Lian;Xingtian Xiao;Jiliang Zhang;Long Jin;Junzhi Yu;Zhongbo Sun","doi":"10.1109/JAS.2024.124425","DOIUrl":"https://doi.org/10.1109/JAS.2024.124425","url":null,"abstract":"This paper presents a distributed scheme with limited communications, aiming to achieve cooperative motion control for multiple omnidirectional mobile manipulators (MOMMs). The proposed scheme extends the existing single-agent motion control to cater to scenarios involving the cooperative operation of MOMMs. Specifically, squeeze-free cooperative load transportation is achieved for the end-effectors of MOMMs by incorporating cooperative repetitive motion planning (CRMP), while guiding each individual to desired poses. Then, the distributed scheme is formulated as a time-varying quadratic programming (QP) and solved online utilizing a noise-tolerant zeroing neural network (NTZNN). Theoretical analysis shows that the NTZNN model converges globally to the optimal solution of QP in the presence of noise. Finally, the effectiveness of the control design is demon-strated by numerical simulations and physical platform experiments.","PeriodicalId":54230,"journal":{"name":"Ieee-Caa Journal of Automatica Sinica","volume":null,"pages":null},"PeriodicalIF":11.8,"publicationDate":"2024-06-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141315195","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
It is interesting yet nontrivial to achieve given control precision within user-assignable time for uncertain nonlinear systems. The underlying problem becomes even more challenging if the transient behavior also needs to be accommodated and only system output is available for feedback. Several key design innovations are proposed to circumvent the aforementioned technical difficulties, including the employment of state estimation fllters with event-triggered mechanism, the construction of a novel performance scaling function and an error transformation. In contrast to most existing performance based works where the stability is contingent on initial conditions and the maximum allowable steady-state tracking precision can only be guaranteed at some unknown (theoretically infinite) time, in this work the output of the system is ensured to synchronize with the desired trajectory with arbitrarily pre-assignable convergence rate and arbitrarily pre-specified precision within prescribed time, using output only with lower cost of sensing and communication. In addition, all the closed-loop signals are ensured to be globally uniformly bounded under the proposed control method. The merits of the designed control scheme are confirmed by numerical simulation on a ship model.
{"title":"Achieving Given Precision Within Prescribed Time yet With Guaranteed Transient Behavior via Output Based Event-Triggered Control","authors":"Zeqiang Li;Yujuan Wang;Yongduan Song","doi":"10.1109/JAS.2023.124134","DOIUrl":"https://doi.org/10.1109/JAS.2023.124134","url":null,"abstract":"It is interesting yet nontrivial to achieve given control precision within user-assignable time for uncertain nonlinear systems. The underlying problem becomes even more challenging if the transient behavior also needs to be accommodated and only system output is available for feedback. Several key design innovations are proposed to circumvent the aforementioned technical difficulties, including the employment of state estimation fllters with event-triggered mechanism, the construction of a novel performance scaling function and an error transformation. In contrast to most existing performance based works where the stability is contingent on initial conditions and the maximum allowable steady-state tracking precision can only be guaranteed at some unknown (theoretically infinite) time, in this work the output of the system is ensured to synchronize with the desired trajectory with arbitrarily pre-assignable convergence rate and arbitrarily pre-specified precision within prescribed time, using output only with lower cost of sensing and communication. In addition, all the closed-loop signals are ensured to be globally uniformly bounded under the proposed control method. The merits of the designed control scheme are confirmed by numerical simulation on a ship model.","PeriodicalId":54230,"journal":{"name":"Ieee-Caa Journal of Automatica Sinica","volume":null,"pages":null},"PeriodicalIF":15.3,"publicationDate":"2024-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142137594","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Dear Editor, This letter considers a collision-free trajectory tracking problem for performance-guaranteed mobile robots (MRs) subject to obstacles. We propose a safety-critical performance-guaranteed trajectory tracking method based on control barrier functions (CBFs). First, an auxiliary system is established to generate the non-negative signals for inflexible bounds such that the performance constraints are not violated when avoiding obstacles. Next, the desired guidance laws are devised to evolve tracking errors within performance space by the error transformation technique. Then, a position-heading CBF based on a heading collision-free principle is developed. Under the CBF framework, the safety-critical angle speed guidance law is solved by a quadratic program with respect to position-heading CBF constraints. It is proved that all errors can converge and evolve within a prescribed performance space, and the closed-loop system is ensured to be safe. Finally, simulation and experiment results are given to verify the effectiveness and feasibility of the proposed control scheme.
{"title":"Safety-Critical Trajectory Tracking for Mobile Robots with Guaranteed Performance","authors":"Wentao Wu;Di Wu;Yibo Zhang;Shukang Chen;Weidong Zhang","doi":"10.1109/JAS.2023.123864","DOIUrl":"https://doi.org/10.1109/JAS.2023.123864","url":null,"abstract":"Dear Editor, This letter considers a collision-free trajectory tracking problem for performance-guaranteed mobile robots (MRs) subject to obstacles. We propose a safety-critical performance-guaranteed trajectory tracking method based on control barrier functions (CBFs). First, an auxiliary system is established to generate the non-negative signals for inflexible bounds such that the performance constraints are not violated when avoiding obstacles. Next, the desired guidance laws are devised to evolve tracking errors within performance space by the error transformation technique. Then, a position-heading CBF based on a heading collision-free principle is developed. Under the CBF framework, the safety-critical angle speed guidance law is solved by a quadratic program with respect to position-heading CBF constraints. It is proved that all errors can converge and evolve within a prescribed performance space, and the closed-loop system is ensured to be safe. Finally, simulation and experiment results are given to verify the effectiveness and feasibility of the proposed control scheme.","PeriodicalId":54230,"journal":{"name":"Ieee-Caa Journal of Automatica Sinica","volume":null,"pages":null},"PeriodicalIF":15.3,"publicationDate":"2024-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10551317","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141991471","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Dear Editor, This letter presents a distributed adaptive second-order latent factor (DAS) model for addressing the issue of high-dimensional and incomplete data representation. Compared with first-order optimizers, a second-order optimizer has stronger ability in approaching a better solution when dealing with the non-convex optimization problems, thus obtaining better performance in extracting the latent factors (LFs) well representing the known information from high-dimensional and incomplete data. However, a traditional second-order optimizer are inefficient in exploiting the curvature information of an LF model due to its large number of parameters. In order to reduce the computational overhead, an inexact second-order method relying on the Hessian-free optimization is preferred. However, this method requires careful coordination of its components, which is time-consuming and impractical for real applications. To address the above issues, the DAS model leverages the curvature information with a Hessian-vector-incorporated inexact second-order optimizer and embeds it into a distributed, multi-phase, and multi-elitist learning particle swarm optimization (DM2PSO) framework for efficient hyper-parameters adaptation and performance gain. Experimental results demonstrate that the DAS model outperforms several state-of-the-art models in estimating missing data on several high-dimensional and incomplete datasets from real-world applications.
{"title":"A Distributed Adaptive Second-Order Latent Factor Analysis Model","authors":"Jialiang Wang;Weiling Li;Xin Luo","doi":"10.1109/JAS.2024.124371","DOIUrl":"https://doi.org/10.1109/JAS.2024.124371","url":null,"abstract":"Dear Editor, This letter presents a distributed adaptive second-order latent factor (DAS) model for addressing the issue of high-dimensional and incomplete data representation. Compared with first-order optimizers, a second-order optimizer has stronger ability in approaching a better solution when dealing with the non-convex optimization problems, thus obtaining better performance in extracting the latent factors (LFs) well representing the known information from high-dimensional and incomplete data. However, a traditional second-order optimizer are inefficient in exploiting the curvature information of an LF model due to its large number of parameters. In order to reduce the computational overhead, an inexact second-order method relying on the Hessian-free optimization is preferred. However, this method requires careful coordination of its components, which is time-consuming and impractical for real applications. To address the above issues, the DAS model leverages the curvature information with a Hessian-vector-incorporated inexact second-order optimizer and embeds it into a distributed, multi-phase, and multi-elitist learning particle swarm optimization (DM2PSO) framework for efficient hyper-parameters adaptation and performance gain. Experimental results demonstrate that the DAS model outperforms several state-of-the-art models in estimating missing data on several high-dimensional and incomplete datasets from real-world applications.","PeriodicalId":54230,"journal":{"name":"Ieee-Caa Journal of Automatica Sinica","volume":null,"pages":null},"PeriodicalIF":15.3,"publicationDate":"2024-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10551316","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142397460","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Autonomous navigation for intelligent mobile robots has gained significant attention, with a focus on enabling robots to generate reliable policies based on maintenance of spatial memory. In this paper, we propose a learning-based visual navigation pipeline that uses topological maps as memory configurations. We introduce a unique online topology construction approach that fuses odometry pose estimation and perceptual similarity estimation. This tackles the issues of topological node redundancy and incorrect edge connections, which stem from the distribution gap between the spatial and perceptual domains. Furthermore, we propose a differentiable graph extraction structure, the topology multi-factor transformer (TMFT). This structure utilizes graph neural networks to integrate global memory and incorporates a multi-factor attention mechanism to underscore elements closely related to relevant target cues for policy generation. Results from photorealistic simulations on image-goal navigation tasks highlight the superior navigation performance of our proposed pipeline compared to existing memory structures. Comprehensive validation through behavior visualization, interpretability tests, and real-world deployment further underscore the adaptability and efficacy of our method.
{"title":"Cognitive Navigation for Intelligent Mobile Robots: A Learning-Based Approach with Topological Memory Configuration","authors":"Qiming Liu;Xinru Cui;Zhe Liu;Hesheng Wang","doi":"10.1109/JAS.2024.124332","DOIUrl":"https://doi.org/10.1109/JAS.2024.124332","url":null,"abstract":"Autonomous navigation for intelligent mobile robots has gained significant attention, with a focus on enabling robots to generate reliable policies based on maintenance of spatial memory. In this paper, we propose a learning-based visual navigation pipeline that uses topological maps as memory configurations. We introduce a unique online topology construction approach that fuses odometry pose estimation and perceptual similarity estimation. This tackles the issues of topological node redundancy and incorrect edge connections, which stem from the distribution gap between the spatial and perceptual domains. Furthermore, we propose a differentiable graph extraction structure, the topology multi-factor transformer (TMFT). This structure utilizes graph neural networks to integrate global memory and incorporates a multi-factor attention mechanism to underscore elements closely related to relevant target cues for policy generation. Results from photorealistic simulations on image-goal navigation tasks highlight the superior navigation performance of our proposed pipeline compared to existing memory structures. Comprehensive validation through behavior visualization, interpretability tests, and real-world deployment further underscore the adaptability and efficacy of our method.","PeriodicalId":54230,"journal":{"name":"Ieee-Caa Journal of Automatica Sinica","volume":null,"pages":null},"PeriodicalIF":15.3,"publicationDate":"2024-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141991449","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","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：