【作者】 张强；

【导师】 吴庆宪；

【作者基本信息】 南京航空航天大学 ， 控制理论与控制工程， 2013， 博士

【摘要】 近空间飞行器(Near Space Vehicle,NSV)集传统航空器与航天器的优点于一身，在军事和民用方面具备重大的战略价值，已成为各军事强国关注和研究的焦点。NSV飞行环境的特殊性使其具有复杂非线性、快时变、强耦合的对象特征，并且在飞行过程中呈现多飞行状态、多任务模式的特点，这使得NSV飞行控制系统的设计更加复杂。围绕这一基础科学问题，本文基于6自由度十二状态近空间飞行器数学模型，在不确定环境下飞行姿态、协调转弯和纵向飞行等方面开展了深入的研究，主要研究成果有：首先，根据国外公开发表的文献资料、相关研究报告与实验室前期研究成果，建立6DOF十二状态NSV运动方程。为解决多变量、强耦合和复杂非线性的NSV姿态控制，研究了基于backstepping方法的一类多输入多输出非线性系统控制器设计方法，并以此设计了NSV姿态角控制器和角速度控制器，仿真结果说明所提方法有效性。其次，围绕具有不确定和外部干扰的NSV姿态运动方程，提出了基于干扰估计的backstepping的控制方法。为补偿系统中的未知干扰，给出两种在线估计算法—自适应干扰估计算法和二阶滑模干扰观测器，并通过理论推导和仿真分析验证了所提出的控制策略能够有效地提高受扰NSV姿态控制系统的抗干扰能力。随后，围绕具有全包线飞行特点的变后掠翼NSV，采用切换仿射非线性系统来描述其不同阶段姿态模型。当切换系统精确已知时，依据backstepping方法，研究一类切换非线性系统光滑控制器设计新方法，以保证模型切换瞬间控制量的光滑性。当切换系统存在未知干扰时，为准确地估计各模态的未知干扰，设计一组径向基神经网络干扰观测器，并采用鲁棒项使得干扰估计值以任意小的误差逼近干扰真值。在此基础上，给出不确定切换非线性系统的控制器结构，并采用多Lyapunov函数推导闭环切换非线性系统稳定条件。以此设计后掠翼不确定时NSV的姿态角控制器和角速度控制器，仿真结果说明所提方法对不确定和外部干扰表现出良好的鲁棒性。然后，围绕NSV纵向姿态/轨迹飞行协调控制问题，提出基于快速收敛非线性微分器干扰补偿的纵向姿态/轨迹飞行协调控制方案。根据NSV纵向飞行特点，给出了NSV纵向姿态/轨迹协调控制模型。针对其中的非仿射非线性航迹模型，提出一种在线的非仿射非线性系统近似方法。为估计NSV纵向运动中的未知复合干扰，给出一种快速收敛的非线性微分器设计方法，并通过理论推导证明其微分估计误差可以收敛于任意小的有界集内。在此基础上，依据受限指令滤波backstepping方法和动态逆方法，设计受扰NSV姿态/轨迹鲁棒协调控制器，并通过理论推导和仿真分析说明所提出的控制策略能够实现姿态与轨迹之间的协调运动。最后，围绕NSV协调转弯控制问题，提出基于二阶滑模干扰观测器干扰补偿的NSV协调转弯鲁棒控制策略。根据NSV转弯飞行特点，建立航迹与姿态角和角速度之间相互影响的NSV协调转弯控制模型。将动态逆方法、带有调节因子的滑模控制方法和二阶滑模干扰观测器相结合，给出重心轨迹协调转弯鲁棒控制器设计过程，其中利用带有调节因子的滑模控制方法以改善航迹倾斜角和航迹方位角的动态性能，进而协调纵向与横侧向运动。仿真结果表明该方法能够有效地完成受扰NSV协调转弯控制任务。更多还原

【Abstract】 Near space vehicle (NSV) has great strategic value both in military and civilian area, which hasboth advantages of traditional aircrafts and spacecrafts. Due to the characteristics of multi-task modesin different phases of flight, NSV possesses some distinct characters, such as serious nonlinearity, fasttime variation, and intense coupling. Therefore, it is a great challenge to design flight control systemsfor NSV. For these problems, some works are studied in this dissertation by using NSV mathematicalmodel of six degrees of freedom and twelve states, such as the flight attitude control, longitudinalflight control and coordinated turn in uncertain environment. The main results of the subject areexhibited as follows:First of all, the6DOF(six degrees of freedom) twelve states equations of NSV are established fordescribing NSV dynamic system, based on the research contributions of the abroad institution and ourlaboratory. In order to solve the problem of NSV attitude control with the multi-variables, strongcoupling and complex nonlinearity, we propose a backstepping scheme for a class of multi-inputmulti-output nonlinear systems. Thus, NSV attitude angle controller and angular velocity controllerare presented by using the proposed method. Simulation results show the desired tracking controlperformance is achieved by the proposed method.Secondly, a backstepping control method based on the disturbance estimation is proposed for theNSV attitude system with uncertainty and external disturbances. Two types of online estimationalgorithms including an adaptive interference estimation algorithm and a second order sliding modedisturbance observer, are given to compensate the unknown disturbances. The good performance ofthis method is demonstrated by both of the theoretical analysis and simulation results.Next, the switched nonlinear systems are used to describe the NSV attitude model of differentstages, when the NSV’s wing sweep angle is variable in large envelop. A smooth backstepping controlmethod is presented for a class of the precisely known switched nonlinear systems, to solve theproblem of controlled variables jumping caused by mode switching. To accurately estimate unknowndisturbances, we design a set of radial basis function neural network disturbance observers, and usesome robust terms to offset the impact on system performance caused by the approximation errors.Then, we derive the structure of uncertain switched nonlinear controller, and analyze the closed-loopstability condition by using the multiple Lyapunov functional method. Based on the above works, anattitude angle controller and an angular velocity controller are designed for the uncertain NSV model. Simulation results show the desired robustness of the proposed method for uncertainty and externaldisturbances.Subsequently, a longitudinal attitude-trajectory coordinate control scheme based on fastconvergence of nonlinear differential disturbance compensator is proposed, aiming at the issue of thelongitudinal coordinate control. According to the characteristics of NSV longitudinal flight, weintroduce NSV longitudinal attitude-trajectory coordinate control equations. A novel non-affinenonlinear approximation method is proposed for non-affine nonlinear flight path dynamic model. Inorder to estimate compound disturbances during the flight phase of longitudinal NSV, we present arapidly convergent nonlinear differentiator and prove that differential estimation residuals converge toan arbitrarily small neighborhood. Based on the above works, a NSV attitude-trajectory coordinatecontroller is designed by using the constrained command filter backstepping method and the dynamicinversion method. Simulation results show that the proposed control strategy can achieveattitude-trajectory coordinate movements.Finally, a robust coordinated-turn control strategy based on second order sliding modedisturbance observer is presented, aiming at the issue of realizing NSV coordinated-turn. According tothe characteristics of NSV turning flight, we establish NSV coordinated-turn control model, in whichthe control of flight path, attitude angle and angular velocity can be considered at the same time.Combining the dynamic inversion method, sliding mode control with regulatory factors and slidingmode disturbance observer, we design a robust controller for the gravity center trajectorycoordinated-turn. Sliding mode control with regulatory factors is used to improve dynamicperformance of flight-path angle and heading angle, and to coordinate the longitudinal and lateralmovement. Simulation results show that this method can effectively implement the uncertain NSVcoordinated-turn control task.更多还原