• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文学位 >Graph theoretic framework based cooperative control and estimation of multiple UAVs for target tracking.
【24h】

Graph theoretic framework based cooperative control and estimation of multiple UAVs for target tracking.

机译：基于图论框架的协同控制和多个用于目标跟踪的无人机的估计。

获取原文
获取原文并翻译 | 示例
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Designing the control technique for nonlinear dynamic systems is a significant challenge. Approaches to designing a nonlinear controller are studied and an extensive study on backstepping based technique is performed in this research with the purpose of tracking a moving target autonomously. Our main motivation is to explore the controller for cooperative and coordinating unmanned vehicles in a target tracking application.;To start with, a general theoretical framework for target tracking is studied and a controller in three dimensional environment for a single UAV is designed. This research is primarily focused on finding a generalized method which can be applied to track almost any reference trajectory. The backstepping technique is employed to derive the controller for a simplified UAV kinematic model. This controller can compute three autopilot modes i.e. velocity, ground heading (or course angle), and flight path angle for tracking the unmanned vehicle. Numerical implementation is performed in MATLAB with the assumption of having perfect and full state information of the target to investigate the accuracy of the proposed controller. This controller is then frozen for the multi-vehicle problem.;Distributed or decentralized cooperative control is discussed in the context of multi-agent systems. A consensus based cooperative control is studied; such consensus based control problem can be viewed from the algebraic graph theory concepts. The communication structure between the UAVs is represented by the dynamic graph where UAVs are represented by the nodes and the communication links are represented by the edges. The previously designed controller is augmented to account for the group to obtain consensus based on their communication. A theoretical development of the controller for the cooperative group of UAVs is presented and the simulation results for different communication topologies are shown. This research also investigates the cases where the communication topology switches to a different topology over particular time instants. Lyapunov analysis is performed to show stability in all cases.;Another important aspect of this dissertation research is to implement the controller for the case, where perfect or full state information is not available. This necessitates the design of an estimator to estimate the system state. A nonlinear estimator, Extended Kalman Filter (EKF) is first developed for target tracking with a single UAV. The uncertainties involved with the measurement model and dynamics model are considered as zero mean Gaussian noises with some known covariances. The measurements of the full state of the target are not available and only the range, elevation, and azimuth angle are available from an onboard seeker sensor. A separate EKF is designed to estimate the UAV's own state where the state measurement is available through on-board sensors. The controller computes the three control commands based on the estimated states of target and its own states. Estimation based control laws is also implemented for colored noise measurement uncertainties, and the controller performance is shown with the simulation results.;The estimation based control approach is then extended for the cooperative target tracking case. The target information is available to the network and a separate estimator is used to estimate target states. All of the UAVs in the network apply the same control law and the only difference is that each UAV updates the commands according to their connection. The simulation is performed for both cases of fixed and time varying communication topology. Monte Carlo simulation is also performed with different sample noises to investigate the performance of the estimator. The proposed technique is shown to be simple and robust to noisy environments.
机译：设计非线性动态系统的控制技术是一项重大挑战。研究了一种设计非线性控制器的方法，并对该技术进行了广泛的研究，其目的是自动跟踪运动目标。我们的主要动机是在目标跟踪应用中探索用于协作和协调无人飞行器的控制器。首先，研究一种用于目标跟踪的通用理论框架，并为单个无人机设计三维环境中的控制器。这项研究主要集中在寻找一种可用于跟踪几乎所有参考轨迹的通用方法。采用反推技术来为简化的无人机运动学模型推导控制器。该控制器可以计算三种自动驾驶模式，即速度，地面航向（或航向角）和飞行路径角，用于跟踪无人驾驶车辆。数值实现是在MATLAB中执行的，假定具有目标的完美状态信息，以研究所提出控制器的精度。然后针对多车辆问题冻结该控制器。在多智能体系统的背景下讨论了分布式或分散式协作控制。研究了基于共识的合作控制；这样的基于共识的控制问题可以从代数图理论的概念中看到。 UAV之间的通信结构由动态图表示，其中UAV由节点表示，通信链路由边缘表示。增强了先前设计的控制器，以说明该组基于他们的沟通获得共识。提出了无人机协同组控制器的理论发展，并给出了不同通信拓扑的仿真结果。这项研究还研究了在特定时刻通信拓扑切换到不同拓扑的情况。进行Lyapunov分析以显示所有情况下的稳定性。本论文研究的另一个重要方面是在无法获得完美或完整状态信息的情况下，为该情况实现控制器。这需要设计一个估计器来估计系统状态。非线性估计器，扩展卡尔曼滤波器（EKF）首先被开发用于通过单个UAV进行目标跟踪。测量模型和动力学模型所涉及的不确定性被认为是具有某些已知协方差的零均值高斯噪声。无法获得目标的完整状态的测量值，并且车载搜寻器传感器仅提供范围，仰角和方位角。一个单独的EKF用于估计无人机的自身状态，通过机载传感器可以进行状态测量。控制器根据目标的估计状态及其自身状态计算三个控制命令。针对彩色噪声测量的不确定性，还实现了基于估计的控制律，并通过仿真结果表明了控制器的性能。目标信息可用于网络，并且单独的估计器用于估计目标状态。网络中的所有UAV都应用相同的控制律，唯一的区别是每个UAV根据其连接更新命令。针对固定和时变通信拓扑的情况都执行了仿真。还使用不同的样本噪声执行了蒙特卡洛模拟，以研究估计器的性能。所提出的技术显示出对嘈杂环境简单且健壮。

著录项

  • 作者

    Ahmed, Mousumi.;

  • 作者单位

    The University of Texas at Arlington.;

  • 授予单位 The University of Texas at Arlington.;
  • 学科 Engineering Aerospace.;Engineering Mechanical.
  • 学位 Ph.D.
  • 年度 2012
  • 页码 158 p.
  • 总页数 158
  • 原文格式 PDF
  • 正文语种 eng
  • 中图分类
  • 关键词

  • 入库时间 2022-08-17 11:43:12

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号