【作者】 田永波；

【导师】 陈静；

【作者基本信息】 武汉理工大学 ， 控制理论与控制工程， 2004， 硕士

【摘要】 本课题来源于211项目“土木工程结构振动中心—地震模拟震动台”。 电液伺服地震模拟振动台试验是地震工程重要的研究手段，近十年来得到了迅速的发展。地震模拟振动台试验不仅负载量大，而且能模拟天然地震波和人工地震波对试件加载。由于地震模拟振动台是一个复杂的环境模拟系统，其系统特性随试件环境和试件状态的不同而不同，并呈现出一定的非线性。因此，振动台台面的运动能否准确的再现要求的地震波形是模拟振动台试验的一个关键的问题。为了使振动台台面真实地反映地震环境，必须有一个完善的控制系统。这样的控制系统基本上可分为2种类型：模拟控制和数字控制。但对现有的模拟控制系统来说由于其存在控制精度差、动态范围小等缺点，对精度要求比较高的地震模拟来说，己不能满足要求，而数字控制系统只要选择合适的控制算法，就能有效的解决这些问题。 迭代学习控制是智能控制中具有严格数学描叙的一个分支，控制器在运行过程中不需要辨识系统参数，它特别适合于有着非线性、强耦合、难以建模以及高精度轨迹控制要求的场合。 鉴于此，在分析电液伺服地震模拟振动台的设备构成及其物理机理基础上，针对武汉理工大学土木工程结构振动中心的地震模拟震动台，推导出了电液伺服地震模拟振动台的数学模型。从智能控制的角度出发设计了地震模拟振动台的迭代学习控制算法，并通过计算机仿真验证了其算法的可行性。在实验室现有的硬件平台上设计了地震模拟振动台数控系统的控制软件。该数控系统能有效提高地震模拟振动台波形精度，控制软件界面友好、操作方便。更多还原

【Abstract】 The dissertation is finished based on the fund 211 project ’seismic simulation shaking table’ of Civil Engineering Structure Vibration Center.The test of electro-hydraulic seismic simulation shaking table is a very important research approach for seismic project. It has been developed rapidly in the recent ten years. Shaking table experiment not only has great load factor, but also can simulate crude seismic wave and artificial wave to load tested object. Seismic simulation shaking table is a complex environment simulation system and show non-linearity, which characteristics will be changed with the condition and state of tested object. Thus, it is a key point for shaking table experiment that whether the wave of shaking table can recur requested seismic wave or not. In order to enable recurrence really seismic condition, the shaking table needs a perfect control system. There are two kinds of control system of analog control and digital control. Comparing with the digital control system, the current existing analog control system has much of flaws, such as bad control precision and a small dynamic range etc. And it cannot match good precision request for the control of seismic simulation shaking table. For digital control system, if a good control algorithm can be implemented, the control problem can be solved effectively.The iteration self-leaning control is a rigorous mathematic description branch of intelligent control. The iteration self-leaning controller doesn’t need identify system parameter. It especially is used for the non-linear and great coupling control system, which mathematical model is deduced difficultly and which control precision is requested high.In view of this, the mathematical model of seismic simulation shaking table is deduced based on analyzing constitutes and mechanism of the seismic simulation shaking table in the Civil Engineering Structure Vibration Center of Wuhan University of Technology. The iteration self-learning control algorithm is designed from the angle of intelligent. The result of simulation shows that thealgorithm is feasibility. The control software of the shaking table digital control system is designed in the having been finished hardware platform of electro-hydraulic seismic simulation shaking table. The actual result shows that the digital control system can improve the wave precision of seismic shaking table and control software has a good interface and convenient operation.更多还原