基于磁流变原理的柔性机械臂抑振技术研究

发布时间：2018-09-18 21:37

【摘要】：目前的航天任务成本昂贵、风险较高,航天器发射后不能正确入轨或出现故障将导致巨大的损失。在轨服务技术,既能保证航天器的正常运行,又能减少航天器故障或失效造成的损失,具有巨大的研究价值和应用前景。在轨服务技术涉及柔性机械臂与非合作目标的接触问题。接触目标瞬间,柔性机械臂会受到非合作目标的冲击碰撞力,瞬时碰撞造成空间柔性机械臂基座扰动、产生振动、任务失败。针对这一问题,本论文研究了面向在轨服务的空间柔性机械臂振动控制技术,将磁流变阻尼器引入空间柔性机械臂关节结构中,根据凯恩方法建立N关节6N自由度漂浮基/柔性机械臂的普适动力学方程,并设计了多自由度分布可控阻尼的空间柔性机械臂抑振方法。本论文的研究内容来源于国家自然科学基金资助项目(51305039)、中央高校基本科研业务费专项资金项目(2014PTB-00-01),具体工作如下：首先,建立了N关节6N自由度具有可控阻尼单元的漂浮基/柔性机械臂动力学模型。为实现振动控制,在柔性机械臂关节处引入由磁流变阻尼器和缓冲器组成的可控阻尼单元,将漂浮基座视为一段拓展机械臂杆,并将柔性机械臂离散为由弹簧和阻尼连接的多刚体段,利用凯恩方法进行运动学和动力学分析,得到空间柔性机械臂的动力学方程。分别利用Matlab与Adams软件进行建模,通过对比仿真曲线,验证了所建立的柔性机械臂动力学模型的有效性。其次,基于磁流变原理,研究了磁流变阻尼器的神经网络正逆模型建模、控制方法。针对磁流变阻尼器输出阻尼的非线性及磁滞特性,利用神经网络对复杂模型的辨识能力,对网络拓扑结构、传输函数、性能函数、训练算法进行研究,分别建立了磁流变阻尼器神经网络正逆模型,通过逆正模型串联,实现了对磁流变阻尼器的精确控制。进一步,基于Matlab和Adams软件平台进行联合仿真,验证了方案。最后,基于控制目标,使用微粒群优化算法,设计了多自由度、分布可控阻尼空间柔性机械臂振动控制策略。以所有关节的振动位移最小为准则,设计了目标函数。利用微粒群优化算法,设计控制流程,通过迭代计算得到各关节处的期望最优阻尼力,并控制磁流变阻尼器输出该阻尼力,实现柔性机械臂的振动衰减。同时,设计了基于PID控制器的柔性机械臂振动控制策略。最后以三关节十二自由度柔性机械臂为研究对象,搭建控制系统,分别对微粒群优化控制、PID控制、未控情况进行仿真,通过对比结果,验证空间柔性机械臂微粒群优化控制方案的有效性。
[Abstract]:The current space mission is expensive and risky, and the inaccuracy or failure of the spacecraft after launch will lead to huge losses. On-orbit service technology can not only ensure the normal operation of the spacecraft, but also reduce the loss caused by the failure or failure of the spacecraft. It has great research value and application prospects. The problem of flexible manipulator's contact with non-cooperative targets is studied in this paper. The flexible manipulator will be subjected to the impact force of non-cooperative targets at the moment of contact with the target. The instantaneous impact will cause the disturbance of space flexible manipulator's base, vibration and mission failure. Based on Kane's method, a universal dynamic equation of a 6N-degree-of-freedom floating base/flexible manipulator with N joints is established. A multi-degree-of-freedom distributed controllable damping method for vibration suppression of a flexible space manipulator is designed. Subsidized projects (51305039), the central university basic scientific research business fee special fund project (2014 PTB-00-01), the specific work as follows:
Firstly, a dynamic model of a flexible manipulator with 6N-degree-of-freedom and controllable damping element is established. In order to achieve vibration control, a controllable damping element composed of magnetorheological damper and buffer is introduced at the joint of the flexible manipulator. The floating base is regarded as an extended manipulator rod, and the flexible manipulator is dispersed as a reason. The kinematics and dynamics of the flexible manipulator with spring and damping are analyzed by Kane method, and the dynamic equations of the flexible manipulator are obtained.
Secondly, based on the principle of magnetorheological (MR), the forward and inverse models of MR dampers are modeled and controlled by neural networks. According to the nonlinear and hysteretic characteristics of MR dampers, the network topology, transfer function, performance function and training algorithm are studied by using neural networks to identify complex models. The forward and inverse model of MR damper neural network is established, and the precise control of MR damper is realized by series connection of the inverse model. Furthermore, the scheme is verified by joint simulation based on MATLAB and Adams software platform.
Finally, based on the control objective, a multi-degree-of-freedom, distributed and controllable damping space flexible manipulator vibration control strategy is designed by using particle swarm optimization algorithm. The objective function is designed according to the minimum vibration displacement of all joints. The control flow is designed by using particle swarm optimization algorithm, and the expected maximum of each joint is obtained through iterative calculation. At the same time, the vibration control strategy of the flexible manipulator based on PID controller is designed. Finally, the control system of the three-joint twelve-degree-of-freedom flexible manipulator is set up. Particle swarm optimization control, PID control, uncontrolled control are respectively established. The simulation results verify the effectiveness of the particle swarm optimization control scheme for space flexible manipulator.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TP241;TB535

【参考文献】