The optimization of spectral analysis for signal harmonics

FFT has been applied to technologies of spectrum analysis widely: However, since the signal sampling is;it random, and FFT suffers specific restricts, some defects happen inevitably, which make harmonic parameters cannot show on spectrum exactly, These defects result from frequency scales cannot mat ch with signal characteristics. This paper affords the optimization of spec trum analysis for signal harmonics. By means of scale fine-tuning, the freq uency scales,will match the signal characteristics. The method eliminates b oth of the picket-fence effect and the leakage effect; it makes the harmoni c parameters show on spectrum more accurately, This paper takes reasonable and practical evaluations to verify the theory.