CAREER: Smart Monitoring and Inspection with Unmanned Aerial Vehicles

职业：无人机智能监控和检查

• 批准号: 1350264
• 负责人: Usman Khan
• 金额: $ 44.05万
• 依托单位: Tufts University
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 2014
• 资助国家: 美国
• 起止时间: 2014-02-01 至 2020-01-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1350264&HistoricalAwards=false
• 关键词: CAREER; Smart; Monitoring; Inspection; Unmanned

Wireless Sensor Networks (WSNs) are being envisioned to monitor health and vitals of the nation?s critical infrastructure, e.g. bridges, roads, and buildings. Other applications include monitoring remote space stations and off-earth structures, e.g. the inflatable lunar habitats. A significant challenge in making remote monitoring a reality is data collection from the sensors, as their batteries do not allow long-range transmission. This research enables data collection and inspection via Unmanned Aerial Vehicles (UAVs), invoking subsequent investigation in dynamic task allocation, distributed path planning, and collaborative navigation in GPS-denied/indoor settings. In particular, this study develops the theoretical foundations for autonomous monitoring and inspection, where WSNs operate in a complete harmony with a network of UAVs. The WSN provides intelligible information to the UAVs for responsive and diagnostic actions; while the UAVs coordinate (and may also bring outside directives) to adapt the WSN to the environment and application demands.To undertake this effort, this research casts the predominantly nonlinear UAV control, navigation, and planning problems in a completely linear framework. The linearity of the underlying dynamics builds upon some of the classical work in convex geometry and Euclidean metrics. Subsequently, the distributed control problem is formulated using structured systems theory, by which efficient graph-theoretic controllability and actuation methods are developed. These features play a significant role in establishing rigorous analytical arguments and in implementing a fully functional remote monitoring and inspection prototype. The transformative approach adapts to the application needs and is robust to imperfections in the underlying communication and environment uncertainties. The study further involves several outreach efforts via social coding platforms and workshops for minority, women in engineering, and K-12 students.

无线传感器网络（WSNs）正在设想监测国家的健康和生命体征？关键基础设施，如桥梁、道路和建筑物。其他应用包括监测远程空间站和地球外结构，例如充气月球栖息地。实现远程监控的一个重大挑战是从传感器收集数据，因为它们的电池不允许远距离传输。这项研究使数据收集和检查通过无人机（UAV），调用动态任务分配，分布式路径规划，并在GPS拒绝/室内设置协同导航后续调查。特别是，这项研究开发的理论基础，自主监测和检查，无线传感器网络运行在一个完全和谐的无人机网络。无线传感器网络提供了可理解的信息，无人机的响应和诊断actions.To进行这项工作，本研究投在一个完全线性的框架主要是非线性的无人机控制，导航和规划问题的无人机协调（也可能带来外部指令），以适应无线传感器网络的环境和应用demands. To进行。基本动力学的线性建立在凸几何和欧几里得度量的一些经典工作之上。随后，分布式控制问题，制定使用结构化系统理论，有效的图论可控性和驱动方法。这些功能在建立严格的分析论证和实施功能齐全的远程监测和视察原型方面发挥了重要作用。变革的方法适应应用程序的需求，是强大的基础通信和环境的不确定性的缺陷。该研究还涉及通过社会编码平台和少数民族，工程妇女和K-12学生研讨会的几项外展工作。

项目成果

Distributed stochastic optimization with gradient tracking over strongly-connected networks

• DOI: 10.1109/cdc40024.2019.9029217
• 发表时间: 2019-03
• 期刊: 2019 IEEE 58th Conference on Decision and Control (CDC)
• 作者: Ran Xin;Anit Kumar Sahu;U. Khan;S. Kar

Distributed Nesterov Gradient Methods Over Arbitrary Graphs

任意图上的分布式 Nesterov 梯度方法

• DOI: 10.1109/lsp.2019.2925537
• 发表时间: 2019
• 期刊: IEEE Signal Processing Letters
• 影响因子: 3.9
• 作者: Xin, Ran;Jakovetic, Dusan;Khan, Usman A.
• 通讯作者: Khan, Usman A.

Distributed Heavy-Ball: A Generalization and Acceleration of First-Order Methods With Gradient Tracking

• DOI: 10.1109/tac.2019.2942513
• 发表时间: 2018-08
• 期刊: IEEE Transactions on Automatic Control
• 影响因子: 6.8
• 作者: Ran Xin;U. Khan

Decentralized Stochastic Optimization and Machine Learning: A Unified Variance-Reduction Framework for Robust Performance and Fast Convergence

• DOI: 10.1109/msp.2020.2974267
• 发表时间: 2020-05
• 期刊: IEEE Signal Processing Magazine
• 影响因子: 14.9
• 作者: Ran Xin;S. Kar;U. Khan

Decentralized Optimization Over Time-Varying Directed Graphs With Row and Column-Stochastic Matrices

• DOI: 10.1109/tac.2020.2969721
• 发表时间: 2020-01
• 期刊: IEEE Transactions on Automatic Control
• 影响因子: 6.8
• 作者: Fakhteh Saadatniaki;Ran Xin;Usman A. Khan