Pub Date : 2025-02-11DOI: 10.1109/tcyb.2025.3534268
Guoliang Wang, Yaqiang Lyu, Guangxing Guo
{"title":"Sampled-Data Stochastic Stabilization of Markovian Jump Systems via an Optimizing Mode-Separation Method","authors":"Guoliang Wang, Yaqiang Lyu, Guangxing Guo","doi":"10.1109/tcyb.2025.3534268","DOIUrl":"https://doi.org/10.1109/tcyb.2025.3534268","url":null,"abstract":"","PeriodicalId":13112,"journal":{"name":"IEEE Transactions on Cybernetics","volume":"6 1","pages":""},"PeriodicalIF":11.8,"publicationDate":"2025-02-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143393000","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}
Pub Date : 2025-02-11DOI: 10.1109/tcyb.2025.3534463
Feisheng Yang, Zhenyu Gong, Qinglai Wei, Yifei Lei
{"title":"Secure Containment Control for Multi-UAV Systems by Fixed-Time Convergent Reinforcement Learning","authors":"Feisheng Yang, Zhenyu Gong, Qinglai Wei, Yifei Lei","doi":"10.1109/tcyb.2025.3534463","DOIUrl":"https://doi.org/10.1109/tcyb.2025.3534463","url":null,"abstract":"","PeriodicalId":13112,"journal":{"name":"IEEE Transactions on Cybernetics","volume":"22 1","pages":""},"PeriodicalIF":11.8,"publicationDate":"2025-02-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143393023","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}
Pub Date : 2025-02-10DOI: 10.1109/TCYB.2025.3533035
Nelson Ma;Junyu Xuan;Guangquan Zhang;Jie Lu
Meta-reinforcement learning (meta-RL) algorithms extract task information from experienced context in order to reason about new tasks, and facilitate rapid adaptation. The quality of these contextual representations (or embeddings) is therefore crucial for a meta-RL agent to make effective decisions in unknown environments. Current methods predominantly assume the existence of a single underlying task, but using a single contextual embedding may not be expressive enough to fully capture the broader distribution of task variations that an agent might encounter. Decomposing that information into different representations can allow them to capture more relevant features in context space while applying additional structure that aids downstream exploitation. In this article, we develop global-local embeddings for contextual meta-RL (GLOBEX), an off-policy contextual meta-RL algorithm that decomposes the contextual representation into separate global and local embeddings. The learning process maximizes information retained by the embeddings and utilizes a mutual information constraint to encourage decoupling. Illustrative examples show that our method effectively adapts by identifying global task dynamics and exploiting temporally local signals. In addition, GLOBEX outperforms existing state-of-the-art meta-RL algorithms on standard MuJoCo benchmarks.
{"title":"Global–Local Decomposition of Contextual Representations in Meta-Reinforcement Learning","authors":"Nelson Ma;Junyu Xuan;Guangquan Zhang;Jie Lu","doi":"10.1109/TCYB.2025.3533035","DOIUrl":"10.1109/TCYB.2025.3533035","url":null,"abstract":"Meta-reinforcement learning (meta-RL) algorithms extract task information from experienced context in order to reason about new tasks, and facilitate rapid adaptation. The quality of these contextual representations (or embeddings) is therefore crucial for a meta-RL agent to make effective decisions in unknown environments. Current methods predominantly assume the existence of a single underlying task, but using a single contextual embedding may not be expressive enough to fully capture the broader distribution of task variations that an agent might encounter. Decomposing that information into different representations can allow them to capture more relevant features in context space while applying additional structure that aids downstream exploitation. In this article, we develop global-local embeddings for contextual meta-RL (GLOBEX), an off-policy contextual meta-RL algorithm that decomposes the contextual representation into separate global and local embeddings. The learning process maximizes information retained by the embeddings and utilizes a mutual information constraint to encourage decoupling. Illustrative examples show that our method effectively adapts by identifying global task dynamics and exploiting temporally local signals. In addition, GLOBEX outperforms existing state-of-the-art meta-RL algorithms on standard MuJoCo benchmarks.","PeriodicalId":13112,"journal":{"name":"IEEE Transactions on Cybernetics","volume":"55 3","pages":"1277-1287"},"PeriodicalIF":9.4,"publicationDate":"2025-02-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143385586","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}
Wastewater treatment processes (WWTPs) are operated under multiple operating conditions. Designing an appropriate optimal control strategy based on the identification of operating conditions is crucial for ensuring the safe operation of WWTPs. To effectively deal with the problem of multiple operating conditions in WWTPs, a data-knowledge-driven multiobjective adaptive optimal control (DK-MAOC) strategy is proposed. First, a fuzzy neural network (FNN) is employed as the prediction model to obtain the concentrations of nitrate and total nitrogen. Then, the operating conditions of WWTPs can be determined. Second, an adaptive objective function (AOF) is proposed to dynamically adjust the weights of operating indices to meet the operational requirements of each operating condition. In particular, the AOF integrates operating requirements and tracking errors to simultaneously consider the feasibility of the controller when solving setpoints. Third, due to the differences in data distribution under each operating condition, real-time data during condition changing is insufficient to accurately predict. A data-knowledge-driven model, incorporating operational knowledge into the FNN-based predictive model, is established to predict the future dynamics of WWTPs. Finally, a collaborative gradient descent algorithm is proposed to simultaneously solve for setpoints and control laws. The effectiveness of the proposed DK-MAOC is tested on the Benchmark Simulation Model No. 1. The experimental results indicate that DK-MAOC can effectively avoid the situation of effluent nitrate nitrogen and total nitrogen exceeding the standards while reducing energy consumption of WWTPs. Therefore, the proposed DK-MAOC can guarantee optimal operation of WWTPs.
{"title":"Data-Knowledge-Driven Multiobjective Adaptive Optimal Control for Wastewater Treatment Processes Under Multiple Operating Conditions","authors":"Hong-Gui Han;Yue Zhang;Hao-Yuan Sun;Zheng Liu;Junfei Qiao","doi":"10.1109/TCYB.2025.3531514","DOIUrl":"10.1109/TCYB.2025.3531514","url":null,"abstract":"Wastewater treatment processes (WWTPs) are operated under multiple operating conditions. Designing an appropriate optimal control strategy based on the identification of operating conditions is crucial for ensuring the safe operation of WWTPs. To effectively deal with the problem of multiple operating conditions in WWTPs, a data-knowledge-driven multiobjective adaptive optimal control (DK-MAOC) strategy is proposed. First, a fuzzy neural network (FNN) is employed as the prediction model to obtain the concentrations of nitrate and total nitrogen. Then, the operating conditions of WWTPs can be determined. Second, an adaptive objective function (AOF) is proposed to dynamically adjust the weights of operating indices to meet the operational requirements of each operating condition. In particular, the AOF integrates operating requirements and tracking errors to simultaneously consider the feasibility of the controller when solving setpoints. Third, due to the differences in data distribution under each operating condition, real-time data during condition changing is insufficient to accurately predict. A data-knowledge-driven model, incorporating operational knowledge into the FNN-based predictive model, is established to predict the future dynamics of WWTPs. Finally, a collaborative gradient descent algorithm is proposed to simultaneously solve for setpoints and control laws. The effectiveness of the proposed DK-MAOC is tested on the Benchmark Simulation Model No. 1. The experimental results indicate that DK-MAOC can effectively avoid the situation of effluent nitrate nitrogen and total nitrogen exceeding the standards while reducing energy consumption of WWTPs. Therefore, the proposed DK-MAOC can guarantee optimal operation of WWTPs.","PeriodicalId":13112,"journal":{"name":"IEEE Transactions on Cybernetics","volume":"55 3","pages":"1056-1069"},"PeriodicalIF":9.4,"publicationDate":"2025-02-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143385588","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}
The invariant ellipsoid methodology focuses on minimizing the invariant/attractive set for a linear control system subjects to bounded external disturbances. In this note, the invariant ellipsoid methodology is adapted to multiagent systems (MASs) by leveraging the generalized homogeneous control. A necessary and sufficient condition for the optimal rejection of external disturbances using a homogeneous control protocol is presented. Compared to linear control protocols, the generalized homogeneous approach yields faster convergence and enhanced accuracy. Theoretical results are validated by the numerical simulations of the multiagent system comprised of unicycle mobile robots (UMRs).
{"title":"Invariant Ellipsoids Method for Homogeneous Leader-Following Consensus Control","authors":"Siyuan Wang;Haibin Duan;Min Li;Andrey Polyakov;Gang Zheng","doi":"10.1109/TCYB.2025.3532779","DOIUrl":"10.1109/TCYB.2025.3532779","url":null,"abstract":"The invariant ellipsoid methodology focuses on minimizing the invariant/attractive set for a linear control system subjects to bounded external disturbances. In this note, the invariant ellipsoid methodology is adapted to multiagent systems (MASs) by leveraging the generalized homogeneous control. A necessary and sufficient condition for the optimal rejection of external disturbances using a homogeneous control protocol is presented. Compared to linear control protocols, the generalized homogeneous approach yields faster convergence and enhanced accuracy. Theoretical results are validated by the numerical simulations of the multiagent system comprised of unicycle mobile robots (UMRs).","PeriodicalId":13112,"journal":{"name":"IEEE Transactions on Cybernetics","volume":"55 3","pages":"1323-1331"},"PeriodicalIF":9.4,"publicationDate":"2025-02-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143385774","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}
In this article, a policy-iteration-based active disturbance rejection control (ADRC) is proposed for uncertain nonlinear systems to achieve real-time output tracking performance, regardless of the specific relative degree of the system. The approach integrates a partial control input generator with a policy-iteration-based reinforcement learning (RL) agent for degree weight adjustment. The partial control input generator includes each ith order partial control input, which is constructed following the ADRC design framework for an ith order system. The RL agent adjusts the degree weights (its actions) to enhance the dominance of the partial control input corresponding to the unknown relative degree through iterative policy refinement. The RL agent is designed to minimize the quadratic reward as the performance index function while enhancing the influence of the partial control input associated with the correct relative degree via the policy iteration procedure. All signals in the closed-loop system (including the time-varying degree weights) ensure semi-global uniformly ultimately boundness using the Lyapunov stability theorem and the affinely quadratically stable property. Consequently, the degree weight adjustments by the RL agent do not affect the closed-loop stability. The proposed method does not require system dynamics, specific relative degree, external disturbances, and other state variable sensing beyond output sensing. The performance of the proposed method was validated via simulations for two different-order uncertain nonlinear systems and experiments using a permanent magnet synchronous motor testbed.
{"title":"Policy-Iteration-Based Active Disturbance Rejection Control for Uncertain Nonlinear Systems With Unknown Relative Degree","authors":"Sesun You;Kwankyun Byeon;Jiwon Seo;Wonhee Kim;Masayoshi Tomizuka","doi":"10.1109/TCYB.2025.3532518","DOIUrl":"10.1109/TCYB.2025.3532518","url":null,"abstract":"In this article, a policy-iteration-based active disturbance rejection control (ADRC) is proposed for uncertain nonlinear systems to achieve real-time output tracking performance, regardless of the specific relative degree of the system. The approach integrates a partial control input generator with a policy-iteration-based reinforcement learning (RL) agent for degree weight adjustment. The partial control input generator includes each ith order partial control input, which is constructed following the ADRC design framework for an ith order system. The RL agent adjusts the degree weights (its actions) to enhance the dominance of the partial control input corresponding to the unknown relative degree through iterative policy refinement. The RL agent is designed to minimize the quadratic reward as the performance index function while enhancing the influence of the partial control input associated with the correct relative degree via the policy iteration procedure. All signals in the closed-loop system (including the time-varying degree weights) ensure semi-global uniformly ultimately boundness using the Lyapunov stability theorem and the affinely quadratically stable property. Consequently, the degree weight adjustments by the RL agent do not affect the closed-loop stability. The proposed method does not require system dynamics, specific relative degree, external disturbances, and other state variable sensing beyond output sensing. The performance of the proposed method was validated via simulations for two different-order uncertain nonlinear systems and experiments using a permanent magnet synchronous motor testbed.","PeriodicalId":13112,"journal":{"name":"IEEE Transactions on Cybernetics","volume":"55 3","pages":"1347-1358"},"PeriodicalIF":9.4,"publicationDate":"2025-02-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143385587","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}
Owing to advancements in deep learning technology, vision transformers (ViTs) have demonstrated impressive performance in various computer vision tasks. Nonetheless, ViTs still face some challenges, such as high computational complexity and the absence of desirable inductive biases. To alleviate these issues, the potential advantages of combining eagle vision with ViTs are explored. A bi-fovea visual interaction (BFVI) structure inspired by the unique physiological and visual characteristics of eagle eyes is introduced. Based on this structural design approach, a novel bi-fovea self-attention (BFSA) mechanism and bi-fovea feedforward network (BFFN) are proposed. These components are employed to mimic the hierarchical and parallel information processing scheme of the biological visual cortex, thereby enabling networks to learn the feature representations of the targets in a coarse-to-fine manner. Furthermore, a bionic eagle vision (BEV) block is designed as the basic building unit based on the BFSA mechanism and the BFFN. By stacking the BEV blocks, a unified and efficient family of pyramid backbone networks called eagle ViTs (EViTs) is developed. Experimental results indicate that the EViTs exhibit highly competitive performance in various computer vision tasks, demonstrating their potential as backbone networks. In terms of computational efficiency and scalability, EViTs show significant advantages compared with other counterparts. The developed code is available at https://github.com/nkusyl/EViT.
{"title":"EViT: An Eagle Vision Transformer With Bi-Fovea Self-Attention","authors":"Yulong Shi;Mingwei Sun;Yongshuai Wang;Jiahao Ma;Zengqiang Chen","doi":"10.1109/TCYB.2025.3532282","DOIUrl":"10.1109/TCYB.2025.3532282","url":null,"abstract":"Owing to advancements in deep learning technology, vision transformers (ViTs) have demonstrated impressive performance in various computer vision tasks. Nonetheless, ViTs still face some challenges, such as high computational complexity and the absence of desirable inductive biases. To alleviate these issues, the potential advantages of combining eagle vision with ViTs are explored. A bi-fovea visual interaction (BFVI) structure inspired by the unique physiological and visual characteristics of eagle eyes is introduced. Based on this structural design approach, a novel bi-fovea self-attention (BFSA) mechanism and bi-fovea feedforward network (BFFN) are proposed. These components are employed to mimic the hierarchical and parallel information processing scheme of the biological visual cortex, thereby enabling networks to learn the feature representations of the targets in a coarse-to-fine manner. Furthermore, a bionic eagle vision (BEV) block is designed as the basic building unit based on the BFSA mechanism and the BFFN. By stacking the BEV blocks, a unified and efficient family of pyramid backbone networks called eagle ViTs (EViTs) is developed. Experimental results indicate that the EViTs exhibit highly competitive performance in various computer vision tasks, demonstrating their potential as backbone networks. In terms of computational efficiency and scalability, EViTs show significant advantages compared with other counterparts. The developed code is available at <uri>https://github.com/nkusyl/EViT</uri>.","PeriodicalId":13112,"journal":{"name":"IEEE Transactions on Cybernetics","volume":"55 3","pages":"1288-1300"},"PeriodicalIF":9.4,"publicationDate":"2025-02-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143258602","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：