基于小生境遗传算法的结构频响函数模型确认  

A structural FRF model validation method based on niching genetic algorithm

作　　者：邓忠民[1] 陈志国[1] 常振国[1] 

机构地区：[1]北京航空航天大学宇航学院,北京100191

出　　处：《振动与冲击》2015年第23期93-98,共6页Journal of Vibration and Shock

基　　金：国家自然科学基金(10972019)

摘　　要：提出了一种基于小生境进化(种群进化)的模型确认方法,采用模式置信准则(Signature Assurance Criterion,SAC)和模式比例因子(Crossing Signature Scale Factor,CSSF)与频响测试约束状态下的固有频率相结合的输出特征作为模型确认的输出响应,以降低输出特征个数并可利用全部频响分析的试验数据。参数修正过程中,采用距离判别方法分析和计算实验与仿真两个数据集之间的统计学差异,并通过逐步迭代、小生境进化达到模型确认的目的。算例仿真结果表明:所提出的基于小生境进化方法可用于结构动力学的频响函数模型确认,相关的频响函数输出特征作为动力学响应输出可以得到较高精度的确认结果。A structural dynamic model validation method based on the dynamic niching was presented to give an accurate validation outcome with niche evolution. The signature assurance criterion（ SAC） and crossing signature scale factor（ CSSF） were used to combine natural frequencies under constraint conditions of frequency response tests. Then,the model＇s output features were taken as output responses for model validation to reduce the number of the output features and use all test data of frequency response analysis. The distance discrimination analysis（ DDA） was used to quantify the fitness of each individual in the niche. The objective of model validation was achieved with an iterative test-analysis correlation（ TAC） procedure and the niche mirror evolution. Simulation results showed that the presented method and using SAC and CSSF as the outputs of dynamic response can get satisfactory results in the model validation of structural dynamics.

关 键 词：模型确认 小生境进化 距离判别 频响函数 

分 类 号：V214[航空宇航科学与技术—航空宇航推进理论与工程] TB12[理学—工程力学]