Subsequently , the primary knowledge of sound wave and the model of signal transmission are introduced . at last , the models of the narrowband and broadband array signal processing , the theory of the number of source signal estimation , the theory of invariable estimation and usage of high - order statistics in array signal processing are presented 文中详细介绍了声波传播的基本知识、信号传输模型;窄带和宽带信号的阵列处理模型;信号源个数估计理论;波达方向估计理论以及高阶统计量在阵列信号处理中的应用。
One is to use fourier transformation to convert the source signal from time domain to frequency domain and to discard high frequency harmonious components upwards of 19 ( gb / t14953 - 93 d5 . 3 demanding ) , then to have static huffman coding to the quantized char array which is composed of reserved direct current component and basic wave and each high frequency " s amplitudes and angles . the other is to use discrete wavelet transformation to convert the source signal from time domain to frequency domain and to set the high frequency coefficients that its absolute value is smaller than the given threshold to zero , then to have dynamic huffman coding to the quantized char array which is composed of multiple , wavelet ' s level , datum length , low frequency coefficients and reserved high frequency coefficients . mass simulinks and analyses under the two circumstances have done to show that data compression ratio is small and the relative error is also small and within the permission of engineering and the compression problem can be solved in theory of measured datum of power system 第一种情况的压缩方法为:采用传统的傅立叶变换把原始信号从时间域变换到频率域,舍弃20次及其以上的高次谐波成分(保证了gb / t14953 ? 93d5 . 3要求) ,然后对保留的直流分量、基波和各次谐波的幅值和相角数据量化后和量化时分别乘以的倍数系数构成一个数组,以字符形式保存,采用静态huffman编码对变换数据进行压缩;采用离散小波变换把原始信号从时间域变换到频率域,然后对分解得到的高频系数进行阈值量化处理,对乘以的倍数系数、小波变换的阶数、小波变换后的低频、各级高频以及原始数据长度、量化后的低频系数以及保留的高频系数大小、位置构成一个数组,以字符形式保存,采用动态huffman编码对这个文件进行压缩。
The research results verify that it can improve the monitoring performance of process . 5 ) due to the failure of extracting process features by noise , a process monitoring method based on blind source signal separation with denoising information by wavelet transform is provided 7 )过程信息并非均是独立同分布,对于很多过程,过程信息往往存在着一定的时间结构,有鉴于此,提出了利用过程信息时间结构的过程监控方法,仿真研究表明,这种方法具有比传统ica的方法更好的性能。