多点随机激励频响函数估计

本文以有限富氏变换为基础,讨论了多点随机激励频响函数估计的一般原理和各种可能的激励方案。提出了重叠取样的数据处理方法,以充分利用有限记录所包含的信息,提高估计精度。在微机PDP-11/23系统上编制了可用于多达6点激励的数据采集-数据处理-频响函数估计的功能较强的软件,可用于4种工作方式:(1)连续随机,(2)周期随机,(3)瞬态随机,(4)全相干激励。在自由-自由梁和玻璃钢蜂窝结构直升机尾桨上进行了试验研究。对连续随机激励进行了重叠取样处理,结果表明该方法切实可行。

Multiple Random Input Estimation of Frequency Response Functions

The general theory and practical techniques for multiple random input estimation of FRF are studied based on the finite Fourier transform and least square fitting. A concept is presented taking ovevla.pped data sections when evaluating spectra in order to extract as much as possible the potential information contained in a certain record. A software for up to 6 inputs FRF estimation is designed, running on the mini-computer PDP-11/23 environment. There are 4 input types selectable: (1) .Comtinuous random, (2).Periodic random, (3).Burst random, (4).Full coherent excitation1]. Tests on a free-free beam and a helicopter tail rotor have shown the advantages of the multiple input estimation over the single input one and the effectiveness of the overlapped data sections approach.