实时子结构实验Chang算法的稳定性和精度  

Stability and accuracy of Chang algorithm for real-time substructure testing

作　　者：吴斌[1] 保海娥[1] 

机构地区：[1]哈尔滨工业大学土木工程学院,黑龙江哈尔滨150090

出　　处：《地震工程与工程振动》2006年第2期41-48,共8页Earthquake Engineering and Engineering Dynamics

基　　金：国家自然科学基金项目(50578047;50338020);国家高科技发展项目(2001AA602015)

摘　　要：与慢速拟动力子结构实验相比,实时子结构实验的优点在于它能真实地反映速度相关型试件的特性。实时子结构实验的逐步积分算法通常借用拟动力算法,但是目前液压伺服作动器很难实现速度反馈控制,因而试件速度不能实现原算法的假定值,这样一来算法的稳定性和计算精度将发生改变。台湾学者S.Y.Chang提出一种无条件稳定的显式拟动力算法,本文分析了这种方法应用于实时子结构实验时的稳定性和计算精度。研究发现在实时子结构实验中该方法由无条件稳定变成了有条件稳定的,精度也发生了改变。Compared with the pseudodynamic testing （PDT）, the advantage of real-time substructure testing （RST） is that it can reflect the velocity dependent behaviour of the structural components. It is common practice that pseudodynamic integration algorithm is directly used for RST. But it is difficult to achieve velocity feedback control with current state-of-the-practice of servohydraulic actuator. So the real velocity will not agree with the assumed velocity of the original PDT algorithm, and hence the stability and accuracy of the RST algorithm will be changed. An unconditionally stable explicit pseudodynamic method （ Chang algorithm） was proposed by S. Y. Chang. This paper analyzes the stability and accuracy of Chang algorithm when used in RST. Chang algorithm becomes conditionally stable for RST while it is unconditionally stable for PDT, and also the accuracy of the algorithm changes.

关 键 词：实时子结构实验 拟动力实验 逐步积分算法 稳定性 精度 

分 类 号：P315.972[天文地球—地震学]