论文部分内容阅读
本文在论述多点激励识别频响函数的一般原理和各种激励方式的优缺点的基础上,提出了多点全相干激励识别方法。多点激励识别的关健在于激励力采样矩阵 F要满秩。在随机激励中,以各激励力间的非全相干性来保证矩阵 F 的各行线性无关;在多点全相干激励中,则以安排矩阵 F 的足够列的线性无关来保证 F 满秩。后者因为是确定性激励,在数据采集量,信号处理方式和工作量,噪声抑制,识别精度等各方面都将显示出优越性。以2点激励,6点采集的应用例子验证了这一方法。
Based on the general principle of multi-point excitation recognition frequency response function and the advantages and disadvantages of various excitation methods, this paper proposes a multi-point and all-coherent excitation identification method. The key point of multi-point motivation recognition is that motivation force sampling matrix F should be full rank. In random excitation, the non-total coherence among different excitation forces guarantees the linear independence of each row of the matrix F. In multi-point coherent excitation, the F-rank is guaranteed by arranging enough columns of matrix F to be linearly independent. The latter, because it is deterministic, will show superiority in all aspects of data collection, signal processing and workload, noise suppression and recognition accuracy. This method is validated by the application of 2-point excitation and 6-point acquisition.