基于多体系统传递矩阵法和遗传算法的一种物理参数识别方法  

A physical parameter identification method based on transfer matrix method of multibody system and genetic algorithm

作　　者：马蕾[1] 芮筱亭[1] 杨富锋[1] Laith K Abbas 文兵[1] 

机构地区：[1]南京理工大学发射动力学研究所,江苏南京210094

出　　处：《振动工程学报》2011年第6期607-612,共6页Journal of Vibration Engineering

基　　金：国家自然科学基金资助项目(11102089);南京理工大学自主科研专项计划资助项目(2011ZDJH11)

摘　　要：将多体系统传递矩阵法与遗传优化算法相结合,形成了一种新的基于多体系统传递矩阵法和遗传算法的物理参数识别方法(MS-TMM&GA)。应用多体系统传递矩阵法进行动力学建模以及固有振动特性分析。将参数识别问题转化为优化问题,结合遗传算法,对由系统固有频率和增广特征矢量构造的目标函数全局最小值优化求解。给出了通过系统模态参数识别物理参数的计算步骤以及流程图。通过两个数值算例,表明了该方法的可行性及有效性。该方法对多体系统传递矩阵法和遗传算法进行了结合与拓展,无需建立复杂多体系统的总体动力学方程,涉及矩阵阶次低,即可快速获得高精度的优化计算结果。This article proposes the investigation of the physical parameter identification problem of multibody system.Combining Transfer Matrix Method of Mulibody System（MS-TMM） and the optimization technology based on Genetic Algorithms（GA）,a new physical parameter identification method（MS-TMMGA） is introduced.The dynamic model is developed,and natural vibration of multibody system is analyzed by MS-TMM.The identification problem is exchanged for an optimization problem which is formulated as a global minimum solution of the objective function defined by parameters of natural frequency and augmented eigenvectors of multibody system.And the physical parameters are identified.The computational steps and the flow chart of physical parameter identification based on system modal parameters are presented.Two numerical examples are computed and discussed,which illustrate the feasibility and affectivity of the proposed approach.The proposed method（TMM-GA） combines and extends transfer matrix method of multibody system and genetic algorithm,with the advantages that the global dynamic equations of the complex multibody system are not needed and the orders of involved system matrices are decreased greatly,the high precision and efficiently searching optimal results can be obtained.

关 键 词：多体系统传递矩阵法(MS-TMM) 物理参数识别 遗传算法(GA) 固有振动特性 

分 类 号：TB123[理学—工程力学] TP391[理学—力学]