摘要
智能优化算法以其可在输入输出数据不完备且含噪声污染以及在系统刚度、质量等先验信息缺乏的情况下识别结构参数的优点,近些年来被广泛应用于参数识别。然而,智能算法识别结构参数时容易出现早熟收敛和陷入局部最优的问题,从而导致识别结果误差较大。帝国竞争算法(ICA)作为一种新颖的智能优化算法,在结构参数识别中也同样存在这类缺陷。为此提出一种融合粒子群优化算法(PSO)全局最优思想的改进帝国竞争算法,并将其运用于结构模态参数识别中。五个标准测试函数的测试结果表明,改进帝国竞争算法的性能优于传统帝国竞争算法。最后通过不同加噪环境下简支梁结构的算例分析,进一步验证改进算法可以有效克服寻优过程中早熟收敛、误差较大的缺陷,并且具有良好的抗噪性。
In recent years,the intelligent optimization algorithm has been widely used in structural parameteridentifications due to the advantage that it can be applied under the conditions of limited measurement data,noise pollutedsignal,and shortage of prior knowledge of mass,damping,or stiffness.However,intelligent algorithm is prone to prematureconvergence and falling into the local optimum during identifying structural parameters,which leads to inaccurateidentification results.Imperialist competitive algorithm(ICA),as a new intelligent optimization algorithm,also has suchdefects in structural parameter identification.Thus,an improved imperialist competitive algorithm based on the globaloptimum idea of particle swarm optimization(PSO)is proposed in this paper.This algorithm is used to identify the structuralmodal parameters.The test results of five typical test functions show that the performance of the improved imperialistcompetitive algorithm is better than that of the traditional imperialist competitive algorithm.Finally,the numerical exampleof a simply supported beam structure under ambient excitation with different noises verifies that the improved algorithm caneffectively overcome the premature convergence and identification error,and has a good noise-immunity performance.
作者
邵永亮
常军
SHAO Yong-liang;CHANG Jun(School of Civil Engineering, Suzhou University of Science and Technology,Suzhou 215011, Jiangsu China)
出处
《噪声与振动控制》
CSCD
2017年第2期152-157,共6页
Noise and Vibration Control
基金
江苏省自然科学基金资助项目(BK20141180)
江苏省结构工程重点实验室开放课题(DZ1405)
江苏省建设系统科技项目(2015ZD77)
关键词
振动与波
帝国竞争算法
结构模态参数识别
环境激励
脉冲响应信号
粒子群算法
vibration and wave
imperialist competitive algorithm
structural modal parameters identification
ambient excitation
impulse response signal
particle swarm optimization
