Title :
Fast method of estimation of signal parameters using the discrete Fourier transform
Author :
Hanrahan, H.E. ; Valkin, A.D.
Author_Institution :
Dept. of Electr. Eng., Univ. of the Witwatersrand, Johannesburg, South Africa
Abstract :
When the discrete Fourier transform (DFT) is used to estimate the amplitude, frequency, and phase of components in a complex signal, leakage and the picket fence effect occur. These occur for signals with components not harmonically related to a fundamental, which are of interest in induction motors. A method allowing accurate measurements of amplitude and phase of each tone in such signals is proposed. The method, which uses the area theorem of the Fourier transform, is simpler than existing methods with which it is compared. The amplitude can be accurately estimated by simply adding the frequency samples in the DFT magnitude spectrum around the main lobe due to a harmonic. A judicious choice of tapered window is necessary prior to computing the DFT. Summation of the real part of the DFT yields the phase
Keywords :
fast Fourier transforms; parameter estimation; signal processing; DFT magnitude spectrum; Dolph-Chebyshev windows; Kaiser windows; amplitude measurement; area theorem; discrete Fourier transform; frequency samples; multitone signals; phase measurement; signal parameter estimation; Amplitude estimation; Discrete Fourier transforms; Estimation error; Fourier transforms; Frequency estimation; Frequency measurement; Induction motors; Parameter estimation; Phase estimation; Phase measurement;
Conference_Titel :
Communications and Signal Processing, 1989. COMSIG 1989. Proceedings., Southern African Conference on
Conference_Location :
Stellenbosch
Print_ISBN :
0-87942-713-2
DOI :
10.1109/COMSIG.1989.129020