基于前馈补偿的振动台粒子群迭代学习控制算法  

PSO iterative learning control algorithm for shaking table based on feedforward compensation

作　　者：安欣[1,2] 高峰 杨巧玉[1] 杨学山 AN Xin;GAO Feng;YANG Qiaoyu;YANG Xueshan(CEA Key Lab of Earthquake Engineering and Engineering Vibration,Institute of Engineering Mechanics,China Earthquake Administration(CEA),Harbin 150080,China;Heilongjiang Provincial Higher Institution Key Lab of Measurement Control Technology and Instrument,Harbin University of Science and Technology,Harbin 150080,China)

机构地区：[1]中国地震局工程力学研究所中国地震局地震工程与工程振动重点实验室,哈尔滨150080 [2]哈尔滨理工大学测控技术与仪器黑龙江省高校重点实验室,哈尔滨150080

出　　处：《振动与冲击》2022年第1期213-220,共8页Journal of Vibration and Shock

基　　金：中央级公益性研究所基本科研业务费专项(2020B04);国家重点研发计划资助(2017YFC1500805)。

摘　　要：针对电磁式振动台对地震波信号复现精度低及复现过程中迭代次数较多的问题,在建立准确的振动台模型的基础上,提出了基于加速度模型的前馈逆模型补偿方法,主要提高电磁式振动台低频特性。此外,针对迭代学习控制算法在振动台波形复现中收敛速度慢的问题,提出了带遗忘因子的反馈辅助PD型迭代学习算法,并用改进自适应粒子群算法,离线优化控制律参数,达到提高复现精度减少迭代次数的目的。试验结果表明,该方法可在少量次数的迭代过程中有效的提高复现精度。Here,aiming at problems of low recurrent accuracy and more iterations in seismic wave signal reproduction process of electromagnetic shaking table,based on the establishment of correct shaking table model,a feedforward inverse model compensation method based on acceleration model was proposed to mainly improve low-frequency characteristics of electromagnetic shaking table.In addition,aiming at slow convergence speed of the iterative learning control algorithm in waveform reproduction of shaking table,a feedback aided PD iterative learning algorithm with forgetting factor was proposed,and the improved adaptive PSO algorithm was used to optimize the control law’s parameters off-line,improve reproduction accuracy and reduce number of iterations.The test results showed that this method can effectively improve reproduction accuracy with a small number of iterations.

关 键 词：振动台 前馈逆控制 迭代学习 粒子群(PSO) 

分 类 号：P315.9[天文地球—地震学]