Based on the characteristic that damages occurring in a structure are usually local, an improved particle swarm optimization ( PSO) algorithm for structure damage detection was developed here by adding a zero mutation ratio coefficient (ZMRC) into PSO. The numerical simulation of a single damage for a two-story rigid frame showed that the proposed method has many advantages, such as, faster convergence of objective function, better identification accuracy, robust noise immunity and less effect of size of population, compared with PSO. Further, the effect of the weighted coefficients of objective function on identification accuracy was investigated. All these advantages were very meanful for application of the proposed method in complex structure damage detection.%根据“结构损伤发生在局部”这一信息特征,在粒子群算法中引入零变异率系数,提出一种改进的粒子群结构损伤识别方法.两层刚架单损伤数值仿真研究结果表明:与普通粒子群算法相比,所提出的方法具有收敛速度快、识别结果准确、抗噪能力较强及受种群数影响小的特点,这些优点对于将PSO应用于复杂结构的损伤诊断有着重要意义.
展开▼
