内平衡系统多目标最优模型降阶研究  

Study of Optimal Model Reduction with Multi-object for Internal Balanced System

作　　者：李雪峰[1] 王建国[1] 

机构地区：[1]合肥工业大学土木与水利工程学院,合肥230009

出　　处：《系统仿真学报》2009年第8期2135-2137,2189,共4页Journal of System Simulation

摘　　要：土木工程结构模型的自由度数目很大,按照原结构模型设计的控制器阶次非常高,从而导致主动控制和半主动控制实施过程中时滞效应十分突出。首先采用均衡实现技术对结构模型的状态变量重新排序形成内部均衡的模型,然后分别基于系统的可观度和可控度指标与内平衡系统二次型性能指标来进行结构模型降阶,略去对结构动力反应贡献小的状态变量。通过对Highway Bridge的Benchmark模型进行多目标模型降阶计算,得到最优的降阶模型,说明方法的可行性和有效性。Since the models of civil engineering structures usually possess comparatively large DOFs, controller designed directly with original structure model is high order. So the phenomenon of time delay is usually predominant when the active/semi-active control technologies are applied to suppressing the structure dynamic response. The balanced realization truncation was adopted to obtain the internal balanced model firstly, and then removed state variables witch were less contribution to structure dynamic response to obtain the reduced model with controllability/observability and quadratic characteristic indexes of internal balanced system separately. The feasibility and validity of optimal model reduction with multi-object are demonstrated by simulation for highway bridge benchmark model.

关 键 词：可观度/可控度 内平衡系统 模型降阶 均衡实现 

分 类 号：U441.3[建筑科学—桥梁与隧道工程]