操作型旋翼飞行机器人系统建模与控制方法研究
文献类型：学位论文
| 作者 | 杨斌
|
| 学位类别 | 博士 |
| 答辩日期 | 2016-05-30 |
| 授予单位 | 中国科学院沈阳自动化研究所 |
| 导师 | 刘光军 ; 韩建达 |
| 关键词 | 操作型旋翼飞行机器人 旋翼飞行机器人 系统建模 耦合分析 鲁棒非线性控制 |
| 其他题名 | Research on Modeling and Control Method for Rotor Flying Manipulator |
| 学位专业 | 模式识别与智能系统 |
| 中文摘要 | 操作型旋翼飞行机器人是一种新型机器人系统，它由飞行机器人（通常是旋翼飞行机器人）与作业装置（机械臂）共同组成，该系统具有垂直起降、低空低速飞行、悬停的特点，同时由于安装了机械臂使得该机器人系统具有完成空中抓取物体，实现空中装配的能力。该系统能够在3维空间中自由运动，因而极大地扩展了移动机器人的工作空间，使得该系统具有广阔的应用前景。旋翼飞行机器人加装主动作业装置将使系统的自主控制面临新的问题。首先,旋翼飞行机器人是多变量、强耦合、欠驱动的非线性系统，易受外界扰动影响。这主要是其运动机理造成的，旋翼飞行机器人运动动力是依靠高速旋转旋翼生成的气流与机体间产生相对运动形成的。加载主动作业装置将使其与飞行机器人间产生互相作用的耦合力/力矩并改变系统的动力学特性，这将大大增加系统建模和控制的难度。其次，能对外界对象实施作业和精准操作是操作型旋翼飞行机器人的主要优势，作业期间系统同外界环境接触是不可避免的。操作型旋翼飞行机器人接触模式下控制难点包括下面两方面问题。1）飞行机器人接近操作对象时，将改变其周边气流特性并导致空气动力学参数变化同时也将引起操作对象的不确定运动。2）与外界环境接触时，系统将受到位置刚性约束下的外界力/力矩扰动，它将显著影响飞行机器人的运动特性。这些问题的出现使得现有的适用于旋翼飞行机器人的控制方法效果降低或是失效。由于操作型旋翼飞行机器人的研究尚处于研究初期阶段，本文的研究重点是非接触情形下的系统稳定性控制及旋翼飞行器与手臂的协调控制。针对这两个问题，本文研究的具体内容如下。首先，对由旋翼飞行机器人与多关节手臂构成的操作型飞行机器人系统进行精确建模。该系统属于典型的多刚体系统，采用分析力学——欧拉-拉格朗日方程的方法对该多刚体进行建模，结合简化的旋翼空气动力学模型得到了系统的完整动力学方程。该模型将作为高保真的仿真对象，同时对该模型进行相应简化以用于系统控制方法的设计。其次, 对该系统在平衡点附近进行线性化处理，设计全状态的线性LQR控制器并研究系统在该控制器下的控制效果，仿真研究显示线性控制器只能在平衡点附近起到较好的控制效果，当手臂摆动幅度较大时，线性控制器无法使系统镇定。接着对旋翼飞行机器人和操作型旋翼飞行机器人进行比较研究，得到系统的耦合模型。依据该模型对耦合力/力矩进行量化分析，研究耦合力/力矩同旋翼飞行机器人的位姿及其导数，手臂关节变量及其导数，以及二者的惯性张量及系统的重心之间的关系，并将该耦合视为旋翼飞行机器人受到的不确定项，为旋翼飞行机器人设计独立的非线性鲁棒控制器提供依据。由于具有多关节手臂的操作型旋翼飞行机器人动力学模型极为复杂，对整个系统设计整体非线性控制器非常困难。利用手臂和飞行平台之间耦合模型，提出了考虑耦合影响的旋翼飞行机器人自适应鲁棒非线性控制器。该控制器将耦合影响视为同平台参数和手臂参数相关的系统不确定项，通过增加对该项扰动的抑制能力来保证整个系统稳定性。同时该控制器对大质量比（手臂质量与操作型旋翼飞行机器人系统质量比值）的系统也有较好的控制效果。 |
| 英文摘要 | Rotor flying manipulator (RFM), a new kind of robot system being composed of a rotor flying robot (RFR) and a (several) manipulator(s), can possess capabilities of vertical take-off and landing (VTOL), low altitude and low speed flying, hovering and so on, also due to the installation of manipulator makes the RFM system with abilities of grasping objects and assembly in the air. The RFM system can move freely in the three-dimensional space, thus greatly expanding the working space of the mobile robot, which makes the system has a broad application prospects. The autonomous control of the RFR system is also faced with new problems when the active operating device be installed on the body. First, the RFR system is a nonlinear system with the characteristics of multivariate, strong coupling, underactuated input and serious sensitivity to the external disturbance. The reason is the movement mechanism of the RFR system which rely on high-speed rotation of the rotor and produce the relative motion between the airflow and the body, and then resulting the drive power of motion. When equipped the operating device, the coupling effect between the manipulator and the RFR will change the dynamics characteristics, which should bring enormous difficulties to the modeling and control of the RFM system. Second, a major advantage of the RFM can operate or do fine work with external objects, which means the contaction with external environment is inevitable. The control difficulties of RFM under contact case including the following two aspects of problems. 1) While the RFR system is close to the external objects, the movement of airflow around the RFR will change, which will directly lead to the rotor aerodynamic parameters change and may generate the uncertainty motion of target. 2) When the RFM system contacts with others objects, the system will suffer sustained influence from the external force/torque in the the rigid constrains of position, which will bring obvious effect to the motion characteristics of RFR. The above-mentioned problems make the control of existing RFR difficult to be applied to RFM. Due to the RFM related research is still in the initial stage, this paper focus on system stability control and coordination control of RFR and manipulator in the non-contact mode. Aiming at the two problems, the specific contents of this paper are as follows. Firstly, corresponding to the RFR and the multi-joint manipulator, an accurate model is given. The RFM system belonging to the typical multi-rigid body system, the system model is construted using Euler-Lagrange equation of the analysis of the mechanics, then obtained the complete system dynamic equations by combining with the simplified air dynamic mechanis model. This model will be seen as a high fidelity simulation object and can also simplify appropriately the model for system controller design. Secondly, linearize the nonlinear system equation near the equilibrium point, then design the full state LQR linear controller and study the control effect under this controller, the simulation results show the controller can obtain better control effect only in the neighborhood of the equilibrium point, the linear controller can not make the system stabilization when the swing range of manipulator is large. Then, further research corresponding to the dynamic equation of the RFM system, the coupling model of this system is obtained. Based on the coupling model, quantitatively analysis the coupling force and coupling torque, and study the relation of following parameters with coupling items, such as the pose and its derivative of RFR. Joint angle variable and its derivative of manipulator, as welle as the inertia tensor and the center of gravity of them, regard the coupling as the uncertainty effect on the RFR, which provides the basis for the RFR’s separative, nonlinear and robust controller. Due to the dynamic model of RFM with multi-joint is very complex, it is very difficult to design the nonlinear control for the whole system. An independent, adaptive and robust nonlinear controller considering the coupling model with the RFR is proposed. The controller regards the coupling items as the system uncertainties, which relate to the parameters of platform and manipulator, and ensures the stability of the RFM by increasing disturbance rejection capability corresponding to this item. At the same time, the controller has better control performance to the RFM system with high quality ratio manipulator. |
| 语种 | 中文 |
| 产权排序 | 1 |
| 页码 | 115页 |
| 源URL | [http://ir.sia.cn/handle/173321/19669]  |
| 专题 | 沈阳自动化研究所_机器人学研究室 |
| 推荐引用方式 GB/T 7714 | 杨斌. 操作型旋翼飞行机器人系统建模与控制方法研究[D]. 中国科学院沈阳自动化研究所. 2016. |
入库方式： OAI收割
来源：沈阳自动化研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。