卫星相对姿态智能自适应控制及分布式仿真技术研究

发布时间：2018-02-07 14:14

本文关键词： 卫星相对姿态控制卫星编队神经网络滑模变结构控制反步法分布式仿真　出处：《北京理工大学》2016年硕士论文　论文类型：学位论文

【摘要】：卫星相对姿态协同控制是编队飞行中的重要组成部分,本文对卫星编队飞行相对姿态控制方法和分布式仿真技术进行研究。在对单颗卫星姿态模型研究和结合现有科研成果的基础上,研究了卫星编队飞行的相对姿态建模方法和控制律设计方法。针对伴随航天器相对于参考航天器绕飞的条件建立了相对姿态动力学模型,研究了滑模变结构控制、反步法和基于RBF神经网络的自适应控制方法,为航天器编队飞行系统的分析以及相对姿态控制器设计提供了一定的理论依据。另外,针对数学仿真的局限性,设计并开发了由仿真计算机、嵌入式星载控制器、飞轮模拟器等组成的卫星相对姿态控制分布式仿真系统,为编队卫星姿态跟踪算法的仿真验证提供了有效的技术手段。主要研究成果和创新点如下:1)推导了卫星编队飞行的相对姿态动力学误差模型和相对姿态运动学误差模型,该模型描述了卫星编队飞行过程中伴随星本体坐标系相对主星本体坐标系的相对姿态运动情况。研究了编队卫星姿态跟踪情况下的协同控制和稳定性问题,在摄动力干扰存在的情况下,分别采用滑模变结构控制方法和反步法设计了全状态反馈的相对姿态控制律。2)在系统模型不准确的情况下,利用神经网络可以逼近任意复杂的非线性控制系统的特性,研究卫星相对姿态神经网络智能自适应控制算法。通过神经网络的训练学习,能够根据系统的可测状态准确的估计出系统的未知参数。采用滑模变结构与神经网络相结合的方法设计卫星编队飞行相对姿态RBF神经网络滑模控制律,利用神经网络的自学习能力,根据控制量和滑动面实时调节控制参数,在整个控制过程中取得到达速度、减弱系统抖振的平衡。3)设计了卫星编队相对姿态控制动态闭环仿真系统,该仿真系统由软硬件部件模拟器按真实卫星姿态控制系统组成,展示卫星姿态控制系统各部分的工作情况和相互协作关系。作为算法控制效果验证、器件信号的分析与调试、部件故障检测等多种工作的工具,为下一步半实物仿真实验打下坚实的基础。
[Abstract]:Satellite relative attitude cooperative control is an important part of formation flight. In this paper, the relative attitude control method and distributed simulation technology of satellite formation flight are studied. The relative attitude modeling method and control law design method of satellite formation flight are studied. The relative attitude dynamics model of accompanying spacecraft relative to reference spacecraft is established, and sliding mode variable structure control is studied. The backstepping method and the adaptive control method based on RBF neural network provide a theoretical basis for the analysis of spacecraft formation flight system and the design of relative attitude controller. A distributed simulation system for satellite relative attitude control is designed and developed, which is composed of simulation computer, embedded spaceborne controller and flywheel simulator. The main achievements and innovations are as follows: 1) the relative attitude dynamic error model and the relative attitude kinematics error model of satellite formation flying are derived. The model describes the relative attitude motion of the satellite body coordinate system relative to the main satellite body coordinate system during satellite formation flying, and studies the cooperative control and stability of satellite formation attitude tracking. In the presence of perturbation, sliding mode variable structure control method and backstepping method are used to design the full state feedback relative attitude control law. 2) when the system model is not accurate, The intelligent adaptive control algorithm of satellite relative attitude neural network is studied by using neural network to approximate the characteristics of any complex nonlinear control system. The unknown parameters of the system can be estimated accurately according to the measurable state of the system, and the sliding mode control law of relative attitude of satellite formation flying RBF neural network is designed by combining sliding mode variable structure with neural network. By using the self-learning ability of neural network and adjusting the control parameters according to the control quantity and sliding surface in real time, the speed of arrival is obtained in the whole control process. The dynamic closed-loop simulation system for satellite formation relative attitude control is designed. The simulation system is composed of software and hardware components simulator according to the real satellite attitude control system. Display the working situation and cooperative relationship of each part of the satellite attitude control system. As a tool to verify the control effect of the algorithm, analyze and debug the device signal, detect the fault of components, etc., Lay a solid foundation for the next hardware-in-the-loop simulation experiment.
【学位授予单位】：北京理工大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：V448.2

【相似文献】