摘要
作为一种新颖的优化工具,大爆炸算法(Big Bang-Big Crunch optimization,BB-BC)被成功应用于很多复杂优化问题。结构参数识别一直是结构健康监测的核心问题,利用BB-BC算法进行结构参数识别的研究。该方法的基本思想是通过最小化识别模型与实际结构系统响应的误差,从而将参数识别问题转化成一个多峰值非线性非凸的优化问题,并利用BB-BC算法发现系统参数的最优估计。利用BB-BC算法在输入输出数据不完备且噪声污染条件下,同时在没有系统质量、刚度等先验信息的情况下对结构系统进行了参数识别,并与基于遗传算法(GA)、粒子群(PSO)的参数识别方法进行了比较。结果表明:该方法可以成功地应用于结构参数识别,识别效能更优越。
As a novel optimization tool,Big Bang-Big Crunch optimization（BB-BC） has attracted much attention and yielded promising results for solving complex optimization problems.This paper utilizes the BB-BC for structural parameter estimation which plays key role in health monitoring.The purpose of parameter estimation is to establish the mathematical model of a structural system to fit the behavior of real systems via minimizing the discrepancy between computed and measured responses,which could be formulated as a multi-modal and nonlinear optimization problem with high dimension.Some results obtained with this algorithm are presented for the identification of structure under conditions including limited input/output data,noise polluted signals,and no prior knowledge of mass,or stiffness of the system.The proposed method is also compared to the identification method based on GA and PSO.The numerical examples and comparing results show that the BB-BC algorithm can successfully applied in structural parameter estimation and the identification performance is superior.
出处
《江西科学》
2010年第2期135-140,共6页
Jiangxi Science
基金
国家自然科学基金项目(50708076)
教育部留学回国人员科研启动基金项目
关键词
大爆炸优化算法
粒子群优化
遗传算法
参数识别
Big bang-big crunch optimization algorithm
Particle swarm optimization algorithm
Genetic algorithm
Parameter estimation
