Article
Computer Science, Information Systems
Gan Huang, Ping Zhao, Guanglin Zhang
Summary: This article proposes an energy management strategy based on deep reinforcement learning, taking into account battery thermal effects. By formulating the problem as an optimization task, extracting features using GRU, and utilizing double DQN algorithm, significant energy reduction is achieved across different driving cycles.
IEEE INTERNET OF THINGS JOURNAL
(2022)
Article
Engineering, Aerospace
Yuchen Liu, Zhonghe Jin, Lai Teng
Summary: This article proposes a multidirectional rapid orientation method for micronano space robots based on optimization theory, considering the highly coupled dynamic system and nonholonomic constraints. A dynamic model of satellite-manipulators cooperative operation micronano space robot is established using a combination of virtual manipulator and Newton-Euler method. The particle swarm optimization method is used to solve the time and path nodes of the orientation process, with the acceleration continuous cubic polynomial trajectory planning method and dynamics-based feedback linearization controller as examples. Simulations of three typical spatial orientation problems verify the feasibility and rapidity of this method, and the article also presents the structure of the objective function for optimization in different coorientation tasks.
IEEE TRANSACTIONS ON AEROSPACE AND ELECTRONIC SYSTEMS
(2023)
Article
Automation & Control Systems
Chinmay Maheshwari, Sukumar Srikant, Debasish Chatterjee
Summary: This article investigates a constrained optimal control problem for an ensemble of control systems in a centralized setting, where each system must meet given state and control action constraints while sharing a controller. The study provides first-order necessary conditions for optimality in the form of a Pontryagin maximum principle.
IEEE TRANSACTIONS ON AUTOMATIC CONTROL
(2021)
Article
Computer Science, Artificial Intelligence
Wenhao Li, Xiangfeng Wang, Bo Jin, Dijun Luo, Hongyuan Zha
Summary: Reinforcement learning has shown excellent performance in static action spaces, but faces challenges in time-varying composite action spaces. This paper proposes a structured cooperative reinforcement learning algorithm, called SCORE, based on the centralized critic and decentralized actor framework. By utilizing a graph attention network and a hierarchical variational autoencoder, SCORE promotes tight cooperation between heterogeneous sub-actions and overcomes multi-agent credit assignment problem.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2022)
Article
Computer Science, Information Systems
Liangliang Hao, Yunjian Xu
Summary: This article discusses the importance of occupancy status information in energy management and proposes a new semi-supervised learning-based occupancy estimator. It also presents a scheme using data-driven deep reinforcement learning algorithm for real-time scheduling decisions. Simulation results show that this scheme can reduce energy costs without sacrificing occupants' thermal comfort.
IEEE INTERNET OF THINGS JOURNAL
(2023)
Article
Automation & Control Systems
Shiba Biswal, Karthik Elamvazhuthi, Spring Berman
Summary: This article addresses the problem of stabilizing a discrete-time deterministic nonlinear control system to a target invariant measure using time-invariant stochastic feedback laws. It proposes a solution and validation method for constructing feedback control laws that stabilize the system to a target invariant measure at a maximized rate of convergence.
IEEE TRANSACTIONS ON AUTOMATIC CONTROL
(2021)
Article
Computer Science, Information Systems
Tong Liu, Lei Lei, Kan Zheng, Kuan Zhang
Summary: Autonomous vehicles in a platoon learn efficient car-following policies by utilizing deep reinforcement learning and dynamic programming techniques. The proposed algorithm, FH-DDPG-SS, overcomes the limitations of lower sampling and training efficiency through three key ideas. Simulation using real driving data demonstrates the effectiveness of FH-DDPG-SS, with comparisons against benchmark algorithms and demonstrations of platoon safety and string stability.
IEEE INTERNET OF THINGS JOURNAL
(2023)
Article
Automation & Control Systems
Chien-Liang Liu, Chun-Jan Tseng, Tzu-Hsuan Huang, Jhih-Wun Wang
Summary: Parallel machine scheduling is a common setting in manufacturing facilities, involving scheduling jobs on machines to minimize objective functions. This problem is NP-hard and difficult in practice, especially when unexpected events occur. Deep reinforcement learning shows potential in solving large optimization tasks, and PMS problems can be formulated as such tasks. A novel DRL-based PMS method, called DPMS, considers PMS characteristics to design states, rewards, and dispatching rules for a dynamic environment, yielding promising results.
IEEE TRANSACTIONS ON SYSTEMS MAN CYBERNETICS-SYSTEMS
(2023)
Article
Automation & Control Systems
Shu Luo, Linxuan Zhang, Yushun Fan
Summary: In this study, a hierarchical multiagent deep reinforcement learning (DRL)-based real-time scheduling method named HMAPPO is proposed to address the dynamic partial-no-wait multiobjective flexible job shop scheduling problem. The method consists of objective agent, job agent, and machine agent, with various job selection rules and machine assignment rules designed to achieve temporary objectives at each rescheduling point. Extensive numerical experiments have confirmed the effectiveness and superiority of HMAPPO compared to other known dynamic scheduling methods.
IEEE TRANSACTIONS ON AUTOMATION SCIENCE AND ENGINEERING
(2022)
Article
Computer Science, Information Systems
Dezhi Chen, Qi Qi, Zirui Zhuang, Jingyu Wang, Jianxin Liao, Zhu Han
Summary: This study addresses the UAV control problem and proposes a solution called MFTRPO, which combines the MFG method and trust region policy optimization to maximize communication efficiency, ensure fair communication range and network connectivity, and is scalable and adaptive. Through extensive simulations, MFTRPO consistently outperforms other methods in terms of coverage, fairness, and energy consumption.
IEEE INTERNET OF THINGS JOURNAL
(2021)
Article
Automation & Control Systems
Lixiang Zhang, Chen Yang, Yan Yan, Yaoguang Hu
Summary: In this study, a distributed real-time scheduling (DRTS) framework with cloud-edge collaboration is proposed to optimize the weighted tardiness by considering both processing services sequencing and logistics services assignment. Experimental results demonstrate the significant potential of this method for efficient real-time decision-making in cloud manufacturing.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2022)
Article
Energy & Fuels
Kang Wang, Haixin Wang, Junyou Yang, Jiawei Feng, Yunlu Li, Shiyu Zhang, Martin Onyeka Okoye
Summary: This paper proposes an EV cluster scheduling strategy considering real-time electricity prices based on deep reinforcement learning. It effectively addresses the curse of dimensionality problem and the challenges of uncertainty in user demand and electricity price, achieving a win-win situation between the power grid and EV users.
Article
Automation & Control Systems
Mingsheng Fu, Liwei Huang, Ananya Rao, Athirai A. Irissappane, Jie Zhang, Hong Qu
Summary: Deep reinforcement learning (DRL) based recommender systems are suitable for user cold-start problems as they can progressively capture user preferences. However, most existing DRL-based recommender systems are suboptimal because they use the same policy for different users. To address this, we propose a multitask Markov Decision Process framework where each task represents a set of similar users. We use Q-learning to optimize the framework and consider task uncertainty through mutual information. Experiments on real-world datasets demonstrate the effectiveness of our approach.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2023)
Article
Computer Science, Artificial Intelligence
Junwei Ou, Lining Xing, Feng Yao, Mengjun Li, Jimin Lv, Yongming He, Yanjie Song, Jian Wu, Guoting Zhang
Summary: This article proposes a solution to the satellite range scheduling problem by combining deep reinforcement learning with a heuristic scheduling method. The algorithm decomposes the problem into two subproblems: task assignment and single antenna scheduling. The DRL is used to determine the task assignment process, and the heuristic scheduling method is utilized to solve the single antenna scheduling problem quickly. Experimental results demonstrate that this method effectively deals with the satellite range scheduling problem.
SWARM AND EVOLUTIONARY COMPUTATION
(2023)
Article
Engineering, Aerospace
Stephen F. Bush
Summary: This article discusses the application of time-sensitive networks (TSN) in avionics environments, focusing on the scheduling challenges in resource-constrained networks. The similarities and differences between avionics full-duplex switched Ethernet and IEEE 802.1 Time-Sensitive Networking are compared, and a new form of reliability called polymorphic reconfiguration is introduced. The article also addresses the computational complexity of TSN scheduling and the constraints posed by limited resources, and suggests efficient scheduling approaches and future research directions.
IEEE TRANSACTIONS ON AEROSPACE AND ELECTRONIC SYSTEMS
(2022)
Article
Astronomy & Astrophysics
R. C. Wang, Dalin Li, Tianran Sun, Xiaodong Peng, Zhen Yang, J. Q. Wang
Summary: Soft X-ray imaging is used to observe the Earth's magnetosphere. In this study, CT is investigated for application to the SMILE mission, and a new algorithm based on CBCT and 3-D GAN is proposed to supplement 2-D images and improve reconstruction quality. Validation experiments show that the supplemented profiles are more complete and accurate, indicating the potential of this technique for observing magnetosphere structure and variability.
JOURNAL OF GEOPHYSICAL RESEARCH-SPACE PHYSICS
(2023)
Article
Environmental Sciences
Junwang Huang, Shi Shen, Min Zhao, Changxiu Cheng
Summary: This study aimed to quantitatively assess the summer regional outdoor thermal comfort (OTC) from 1981 to 2020 in the Beijing-Tianjin-Hebei (BTH) urban agglomeration. The results showed a worsening OTC trend with spatial and temporal heterogeneity in the BTH region.
Article
Green & Sustainable Science & Technology
Zhixiao Zou, Changxiu Cheng, Shi Shen
Summary: Maize is highly vulnerable to climate change and water scarcity, making it crucial to improve water utilization efficiency. This study assessed the combined effects of meteorological conditions and irrigation levels on maize yields in the Jing-Jin-Ji region from 1993 to 2019. The results revealed that minimum temperature influenced the sowing period, while maximum temperature affected other growth stages. The inflection point for total precipitation during the ear stage of summer maize was 401.42 mm. Higher precipitation levels led to decreased summer maize yield. Drought significantly impacted maize growth during the seedling stage. Higher effective irrigation rates and longer dry spells increased maize yield. The evaluation results are valuable for ensuring food security and advancing towards a water-energy-food nexus.
Article
Green & Sustainable Science & Technology
Yimeng Xu, Yongjuan Xie, Xudong Wu, Yitian Xie, Tianyuan Zhang, Zhixiao Zou, Rongtian Zhang, Zhiqiang Zhang
Summary: China has made a commitment to strictly maintain wetland ecosystem services and the value of wetland ecosystem services in China needs to be measured. This study introduces the concept of cosmic exergy in evaluating wetland ecosystem services and investigates the temporal-spatial variations of China's wetland ecosystem service value from the donor side. Results show that the total wetland ecosystem service value in China has increased with an overall growth rate of 22%, and the ecological quality of wetlands varies in different regions. Marsh wetlands play a vital role in ecosystem service supply, especially in microclimate regulation.
JOURNAL OF CLEANER PRODUCTION
(2023)
Article
Computer Science, Artificial Intelligence
Xian Li, Mingli Ding, Yanfeng Gu, Aleksandra Pizurica
Summary: This paper introduces a unified deep learning framework for joint denoising and classification of high-dimensional images, particularly in the framework of hyperspectral imaging. The proposed joint learning framework substantially improves the classification performance and enhances the denoising results, especially in terms of the semantic content, benefiting from the classification.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Article
Engineering, Electrical & Electronic
Shengyang Li, Xian Sun, Yanfeng Gu, Yixuan Lv, Manqi Zhao, Zhuang Zhou, Weilong Guo, Yuhan Sun, Han Wang, Jian Yang
Summary: Intelligent processing of satellite video focuses on extracting specific information of ground objects and scenes from earth observation videos through intelligent image/video processing technology. This article presents a systematic review and quantitative analysis of the results published over the last decade, intending to further promote the development of various intelligent processing tasks for satellite video. It analyzes the current difficulties, challenges, and methodological systems for each task, and also provides in-depth analysis and summary of publicly available datasets, algorithm performance, and application scenarios.
IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING
(2023)
Article
Geochemistry & Geophysics
Dakuan Du, Yanfeng Gu, Tianzhu Liu, Xian Li
Summary: In this article, a novel convolution and transformer joint network (CTJN) is proposed to address the challenge of high-accuracy spectral reconstruction (SR) in complex scenes. The CTJN utilizes shallow feature extraction modules (SFEMs) and deep feature extraction modules (DFEMs) to explore local spatial features and global spectral features. Additionally, a high-frequency transformer block (HF-TB) is designed to preserve detailed features and a spatial-spectral recalibration block (SSRB) is incorporated to enforce explicit constraints. Experimental results on multiple datasets demonstrate the superior performance of CTJN compared to state-of-the-art methods in both large- and small-scale scenes.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2023)
Article
Geochemistry & Geophysics
Wen Xie, Yanfeng Gu, Tianzhu Liu
Summary: This article proposes a two-stream encoder-decoder network for the single hyperspectral (HS) task, which includes a reflectance estimation subnetwork (RES) and a shading estimation subnetwork (SES). The network introduces three physical losses to enhance performance and ensures the estimated intrinsic components are physically correct. Experimental results demonstrate that the proposed network outperforms current available learning or optimization-based approaches.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2023)
Article
Geochemistry & Geophysics
Zhe Dong, Guoming Gao, Tianzhu Liu, Yanfeng Gu, Xiangrong Zhang
Summary: This article proposes a novel cross-model knowledge distillation framework, named DSCTs, to harness the complementary advantages of both CNNs and transformers. By learning complementary knowledge from the teacher model, our DSCT framework improves the student's segmentation performance without adding trainable parameters.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2023)
Article
Geochemistry & Geophysics
Yanfeng Gu, Zhen Xiao, Xian Li
Summary: This article presents an accurate spatial alignment method for UAV LiDAR strip adjustment in nonurban scenes. A novel point cloud feature descriptor called SSPF is constructed to extract multidimensional nonstructural features that are robust to nonurban point clouds. The proposed method is validated on two nonurban datasets and proves its superiority compared to mainstream strip adjustment methods.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2023)
Article
Geochemistry & Geophysics
Zhe Dong, Tianzhu Liu, Yanfeng Gu
Summary: This article proposes a spatial and semantic consistency self-supervised contrastive learning (SSCCL) framework for remote sensing semantic segmentation tasks. By integrating a consistency branch and an instance branch, the framework can learn robust and informative feature representations in limited annotated scenarios, achieving superior performance compared to state-of-the-art CL methods and ImageNet pretraining.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2023)
Article
Geochemistry & Geophysics
Likun Chen, Yanfeng Gu, Xian Li, Xiangrong Zhang, Baisen Liu
Summary: An article proposes a normalized spatial-spectral supervoxel segmentation method for multispectral point cloud (MPC) data, which can segment MPC without the need for any manual annotation and achieves better performance compared to other methods, as demonstrated in experiments.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2023)
Article
Geochemistry & Geophysics
Qingwang Wang, Cheng Yin, Haochen Song, Tao Shen, Yanfeng Gu
Summary: In this study, a novel uncertainty-guided trustworthy fusion network (UTFNet) is proposed for RGB-T semantic segmentation. The uncertainty of each modality is estimated and used to guide the information fusion, resulting in improved accuracy, robustness, and trustworthiness of the segmentation model.
IEEE GEOSCIENCE AND REMOTE SENSING LETTERS
(2023)
Article
Geochemistry & Geophysics
Bin Guo, Tianzhu Liu, Yanfeng Gu
Summary: In this article, a novel structure-preserving discriminative distribution adaptive MS-HS image collaborative classification method is proposed to improve the classification accuracy of large-scene MS images. The method maximizes the distance between different classes by combining statistical properties and geometric constraints, and adaptively maps multiscale spectral-spatial features of MS-HS images to subspaces for classification. Experimental results on three sets of MS-HS datasets demonstrate the effectiveness of the proposed method in reducing the differences between MS-HS data and achieving better classification results.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2023)
Article
Geochemistry & Geophysics
Yanfeng Gu, Yanyuan Huang, Tianzhu Liu
Summary: Traditional spectral unmixing of satellite hyperspectral images faces challenges of severe spectral mixing and variability caused by external factors. In this study, a novel UAV-satellite spectral unmixing model with intrinsic image decomposition is proposed to address these problems. Experimental results show that the proposed method effectively improves the robustness and accuracy of the unmixing results.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2023)