摘要
针对标准粒子群优化算法易陷入局部最优及多粒子群协同优化算法(MPSCO)仍存在的不足,提出了一种改进的多粒子群协同优化算法,并将其与Newmark常平均加速度法、假设检验相结合,研发出一种适用于框架结构的损伤检测策略。首先,针对结构损伤前的实测加速度响应,运用改进MPSCO算法识别出结构损伤前的层间刚度值;第二,针对结构损伤后的实测加速度响应并运用同样的方法识别出结构损伤后的层间刚度值;第三,比较所识别出的结构损伤前后的层间刚度值并结合假设检验从而完成损伤检测。最后通过一7层钢框架的数值模型和实验室试验验证了本文所提损伤检测策略的可靠性。研究结果表明,本文所提的损伤检测策略能够精确地识别结构损伤同时具有较好的抗噪性能和鲁棒性。
This paper presents an improved multi-particle particle swarm coevolution （MPSC0） algorithm to over- come the local optima in standard particle swarm optimization （PS0） algorithm and the deficiency of MPSC0, whereby a structural damage detection strategy is proposed by integrating the Newmark＇s algorithm and hypothesis testing. Firstly, in terms of measured acceleration responses, the improved MPSC0 algorithm is used to identify the undamaged stiffness. Secondly, acceleration responses in damage or practical states are processed by the above- mentioned method so as to identify the current stiffness. Thirdly, after the damage threshold is determined by the hypothesis testing, damage detection is performed by comparing the stiffness difference between the undamaged and current stiffness. Finally, numerical simulation and frame dynamic testing in laboratory are validated the proposed strategy. The results show that the presented damage detection strategy not only is effective and applicable, but also has good noise-tolerance and robustness.
出处
《地震工程与工程振动》
CSCD
北大核心
2017年第2期108-116,共9页
Earthquake Engineering and Engineering Dynamics
基金
国家"十二五"科技支撑计划(2015BAK14B02-06)~~
关键词
损伤检测
多粒子群协同优化
NEWMARK法
假设检验
damage detection
improved multi-particle swarm coevolution optimization （MPSCO）
Newmark＇s al- gorithm
hypothesis testing
作者简介
巫文君（1984-），女，博士研究生，主要从事组合结构、抗震监测、评价与修复加固研究．E-mail：190521698@qq．com
通讯作者：姜绍飞（1969-），男，教授，博士，主要从事结构健康监测与组合结构研究．E-mail：cejsf@Nu．edu．cn
