运用RSS-QPSO算法识别结构模态参数  

Identification of Structural Modal Parameters Using RSS-QPSO Algorithm

作　　者：崔竞元 杨杰[1] 程琳[1] 徐笑颜 陈诗怡 CUI Jingyuan;YANG Jie;CHENG Lin;XU Xiaoyan;CHEN Shiyi(Institute ofWater Resources and Hydro-electric Engineering,Xi’an University of Technology,Xi’an 710048,China)

机构地区：[1]西安理工大学水利水电学院,西安710048

出　　处：《噪声与振动控制》2020年第6期59-66,共8页Noise and Vibration Control

基　　金：国家自然科学基金青年科学基金资助项目(51809212);陕西省自然科学基础研究计划重点资助项目(2018JZ5010);陕西省水利科技计划资助项目(2018SLKJ-5)。

摘　　要：运用智能优化模态参数识别方法识别多自由度系统模态时,容易出现早熟收敛和陷入局部最优;改进搜索能力算法多需多次迭代保证结果精度;将多模态信号转换为单模态信号的时频分析方法自身存在缺陷。从模态独立性和传统模态参数识别方法出发,提出一种将搜索空间缩减和量子粒子群算法结合(reducing search space with quantum-behaved particle swarm optimization algorithms,RSS-QPSO)的模态参数识别方法。结合数值算例和悬臂梁实验研究基于RSS-QPSO与量子粒子群算法(QPSO)的识别结果;在不同噪声环境下对比了RSS-QPSO与特征系统实现法(ERA)、随机子空间法(SSI)、峰值拾取法(PP)识别结果。研究结果表明:RSS-QPSO能够一定程度上克服早熟收敛和局部最优缺陷,频率和阻尼比识别精度较高,鲁棒性较强;振型识别精度略差,但鲁棒性好。When using the intelligent optimization of modal parameter identification method for the modal identification of multi-DOF systems,it is easy for calculation to fall into a prematurely converge and a local optimum.The improved search-ability algorithm needs multiple iterations to ensure the necessary accuracy of the result.The time-frequency analysis method converting multi-modal signals to single-modal signals has its own defects.Therefore,based on modal independence and traditional modal parameter identification methods,a modal parameter identification method of“reducing search space with quantum-behaved particle swarm optimization”(RSS-QPSO)algorithm is proposed.Combining numerical examples and cantilever beam experiments,the recognition results of RSS-QPSO and QPSO are compared.The recognition results of RSS-QPSO,eigensystem realization algorithm(ERA),stochastic subspace identification(SSI),and peak picking algorithm(PP)are compared under different noise environments.The research results show that RSS-QPSO can overcome premature convergence and local optimization problems in a certain extent;the frequency and damping ratio recognition accuracy of RSS-QPSO is higher and its robustness is higher;the accuracy of mode shape recognition is slightly lower,but its robustness is better.

关 键 词：振动与波 多自由度系统 量子粒子群算法 模态参数识别 缩减搜索空间 

分 类 号：TP206.3[自动化与计算机技术—检测技术与自动化装置]