压电柔性板自适应滤波振动控制研究
发布时间：2018-10-13 12:06
【摘要】：在航空航天领域中，，飞行器和航天器附件等外表面多采用柔性薄壁结构，以减少结构的总质量和能量损耗等。由于薄壁结构自身低刚度、低质量的特性，当航天器结构受到持续扰动激励起系统的大幅值振动时，或者空天飞行器在超音速飞行时其壁板结构在惯性力、弹性力、热载荷以及气动力的耦合作用下产生壁板颤振时，均会引起结构的疲劳与损伤破坏，因此，对结构的颤振与振动控制研究是非常有必要的。本文针对壁板结构颤振现象发生的条件，设计了一种两端固支的压电柔性板结构来模拟壁板结构，并以此为研究对象进行动力学建模仿真以及振动控制研究。 系统的动力学方程是对其进行研究的理论基础。首先采用一种矩形薄板单元对两端固支的柔性板结构进行了有限元建模，为接下来的传感器/驱动器的优化布置以及控制算法的仿真提供了模型。 其次，基于一种改进的H2范数的优化指标进行压电传感器/驱动器的位置优化配置，该指标不仅考虑将控制模态指标最大化，同时将残余模态指标最小化，以保证控制系统具有较好的可控性和可观性。分别采用遗传算法寻优以及枚举法的方式对比验证了传感器/驱动器的最优位置。建立了压电柔性板的动力学模型，进行了压电板的特性分析，并采用ANSYS软件进行对比分析。 再次，分别采用经典PD反馈控制算法、自适应滤波-X LMS(Least Mean Square，最小均方误差)前馈控制算法以及自适应滤波-U LMS前馈控制算法，对持续激励扰动下两端固支压电板的振动控制进行仿真研究。 最后，基于前述的理论研究以及仿真分析结果的基础上，搭建了两端固支压电板的振动主动控制实验平台，进行了通道模型辨识，并对相应的控制算法PD，FXLMS，FULMS进行了实验研究和对比分析。 基于改进的H2范数的优化指标进行压电传感器/驱动器位置优化的可行性得到了验证。仿真和实验研究分析的结果表明，采用PD反馈控制算法以及自适应滤波前馈控制算法均能有效地抑制壁板结构在持续激励下的振动。并且，FXLMS和FULMS前馈控制算法可以把一阶模态持续激励下的振动通过较长时间的调整后，抑制到很小的幅值，相比反馈算法，这就是所研究的前馈算法的优势。
[Abstract]:In the field of aeronautics and astronautics, flexible thin-walled structures are often used on outer surfaces such as aircraft and spacecraft accessories to reduce the total mass and energy loss of the structures. Due to the low stiffness and low mass of thin-walled structures, when the spacecraft structure is subjected to continuous disturbance to excite the large amplitude vibration of the system, or when the spaceflight is flying at supersonic speed, the wall structure of the spacecraft is subjected to inertial force and elastic force. The fatigue and damage of the structure can be caused by the flutter of the wall plate under the coupling of thermal load and aerodynamic force, so it is necessary to study the flutter and vibration control of the structure. In this paper, a piezoelectric flexible plate structure with fixed two ends is designed to simulate the wall plate structure, which is based on the flutter phenomenon of the wall structure. The dynamic modeling and simulation and vibration control research are carried out on the basis of the design of the piezoelectric flexible plate structure. The dynamic equation of the system is the theoretical basis of its research. Firstly, a rectangular thin plate element is used to model the flexible plate structure fixed at both ends by finite element method, which provides a model for the optimization of sensor / driver arrangement and the simulation of control algorithm. Secondly, based on an improved H _ 2 norm optimization index, the piezoelectric sensor / actuator position optimization is carried out, which not only considers maximizing the control modal index, but also minimizes the residual modal index. In order to ensure that the control system has better controllability and observability. The optimal position of sensor / driver is verified by genetic algorithm and enumeration method respectively. The dynamic model of piezoelectric flexible plate is established. The characteristics of piezoelectric plate are analyzed and compared by ANSYS software. Thirdly, the classical PD feedback control algorithm, adaptive filter-X LMS (Least Mean Square, minimum mean square error) feedforward control algorithm and adaptive filter-U LMS feedforward control algorithm are used respectively. The vibration control of piezoelectric plates with two ends clamped under continuous excitation and disturbance is simulated. Finally, based on the theoretical research and simulation results, an experimental platform for active vibration control of two clamped piezoelectric plates is built, and the channel model is identified. The corresponding control algorithm, PD,FXLMS,FULMS, is studied experimentally and compared. The feasibility of piezoelectric sensor / actuator position optimization based on the improved H _ 2 norm optimization index is verified. The results of simulation and experimental analysis show that both the PD feedback control algorithm and the adaptive filter feedforward control algorithm can effectively suppress the vibration of the wall structure under continuous excitation. Moreover, FXLMS and FULMS feedforward control algorithms can suppress the vibration of the first order mode to a very small amplitude after a long period of adjustment, which is the advantage of the feedforward algorithm compared with the feedback algorithm.
【学位授予单位】：华南理工大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TB535;TP13
本文编号：2268557
[Abstract]:In the field of aeronautics and astronautics, flexible thin-walled structures are often used on outer surfaces such as aircraft and spacecraft accessories to reduce the total mass and energy loss of the structures. Due to the low stiffness and low mass of thin-walled structures, when the spacecraft structure is subjected to continuous disturbance to excite the large amplitude vibration of the system, or when the spaceflight is flying at supersonic speed, the wall structure of the spacecraft is subjected to inertial force and elastic force. The fatigue and damage of the structure can be caused by the flutter of the wall plate under the coupling of thermal load and aerodynamic force, so it is necessary to study the flutter and vibration control of the structure. In this paper, a piezoelectric flexible plate structure with fixed two ends is designed to simulate the wall plate structure, which is based on the flutter phenomenon of the wall structure. The dynamic modeling and simulation and vibration control research are carried out on the basis of the design of the piezoelectric flexible plate structure. The dynamic equation of the system is the theoretical basis of its research. Firstly, a rectangular thin plate element is used to model the flexible plate structure fixed at both ends by finite element method, which provides a model for the optimization of sensor / driver arrangement and the simulation of control algorithm. Secondly, based on an improved H _ 2 norm optimization index, the piezoelectric sensor / actuator position optimization is carried out, which not only considers maximizing the control modal index, but also minimizes the residual modal index. In order to ensure that the control system has better controllability and observability. The optimal position of sensor / driver is verified by genetic algorithm and enumeration method respectively. The dynamic model of piezoelectric flexible plate is established. The characteristics of piezoelectric plate are analyzed and compared by ANSYS software. Thirdly, the classical PD feedback control algorithm, adaptive filter-X LMS (Least Mean Square, minimum mean square error) feedforward control algorithm and adaptive filter-U LMS feedforward control algorithm are used respectively. The vibration control of piezoelectric plates with two ends clamped under continuous excitation and disturbance is simulated. Finally, based on the theoretical research and simulation results, an experimental platform for active vibration control of two clamped piezoelectric plates is built, and the channel model is identified. The corresponding control algorithm, PD,FXLMS,FULMS, is studied experimentally and compared. The feasibility of piezoelectric sensor / actuator position optimization based on the improved H _ 2 norm optimization index is verified. The results of simulation and experimental analysis show that both the PD feedback control algorithm and the adaptive filter feedforward control algorithm can effectively suppress the vibration of the wall structure under continuous excitation. Moreover, FXLMS and FULMS feedforward control algorithms can suppress the vibration of the first order mode to a very small amplitude after a long period of adjustment, which is the advantage of the feedforward algorithm compared with the feedback algorithm.
【学位授予单位】：华南理工大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TB535;TP13
【参考文献】
相关期刊论文 前10条
1 苑凯华;邱志平;;压电复合材料壁板颤振的控制[J];北京航空航天大学学报;2009年12期
2 谢建宏;张为公;梁大开;;智能材料结构的研究现状及未来发展[J];材料导报;2006年11期
3 钱振东,黄卫,朱德懋;主动减振板的数值分析[J];工程力学;2000年02期
4 常军,陈敏,刘正兴;基于能量准则的梁振动多模态主动控制的LQR法[J];固体力学学报;2004年04期
5 姚伟岸;杨柳;;压电材料三维问题的虚边界元——最小二乘配点法[J];计算力学学报;2007年06期
6 王忠东,陈塑寰;智能结构有限元动力模型的建立及主动振动控制和抑制[J];计算力学学报;1998年01期
7 李山青,刘正兴,杨耀文;压电材料在智能结构形状和振动控制中的应用[J];力学进展;1999年01期
8 董聪，夏人伟;智能结构设计与控制中的若干核心技术问题[J];力学进展;1996年02期
9 李俊宝,张景绘,任勇生,张令弥;振动工程中智能结构的研究进展[J];力学进展;1999年02期
10 孙东昌,王大钧;智能板振动控制的分布压电单元法[J];力学学报;1996年06期
相关博士学位论文 前1条
1 李双蓓;压电智能结构分析的新方法研究及其应用[D];广西大学;2012年
本文编号：2268557
本文链接：https://www.wllwen.com/guanlilunwen/gongchengguanli/2268557.html
