前言用快速付里叶变换(FFT)方法进行频谱相关分析已经深入到许多工程技术领域,其重要性日益为人们所识认。然而,在实际应用中,还不十分令人满意最突出的问题是分析的频率范围和分辨力之间的矛盾。在工程问题中,常会遇到要观测信号的频率范围很宽,而需要高分辨力进行观测的频带却又很窄的情况。用通常FFT 的方法进行频谱分析,其频率分析范围总是从直流到奈奎斯特频率,并且在这样宽的频率范围内,具有均匀的频率分辨力。就是说,为了在需要进行细微观测的窄带内得到高分辨力频谱。则必须使整个分析频带都具有相同的高分辨力,这是十分浪费的。例如,如果对全频带进行观测,用M 点采样和变换,分辨力就能满足要求,但若 Preface Spectrum correlation analysis using the fast Fourier transform (FFT) method has been extended to many engineering and technical fields, and its importance has become increasingly recognized. However, it is not quite satisfactory in practical application. The most prominent problem is the contradiction between the frequency range and the resolution of the analysis. In engineering problems, it is often the case that the frequency range of the observed signal is very wide and the frequency band in which high-resolution observation is required is very narrow. The spectrum analysis is performed using the usual FFT method, whose frequency analysis always ranges from DC to Nyquist and has uniform frequency resolution over such a wide range of frequencies. That is to say, in order to obtain high-resolution spectrum in a narrow band that requires microscopic observation. It is very wasteful to have the same high resolution for the entire analysis band. For example, if the whole band is observed, the resolution is sufficient by M-point sampling and transformation, but if