Abstract
This paper examines a modified version of the instantaneous power spectrum (IPS). The short duration periodogram and the pseudo Wigner–Ville distribution (PWD) have been used extensively to display the time-varying spectral information. Disadvantages of the periodogram are its inherently poor resolution due to the requirement of matching the resolution width to the spectral dynamics of the signal. Potential problems associated with PWD are spectral cross terms and an inability to resolve spectral components at the endpoints of the analysis segment. A modified version of IPS is shown to be less sensitive to these types of problems. In addition, IPS can work with real valued signals sampled at the traditional Nyquist rate. © 1990 IEEE

Abstract

Abstract
A novel algorithm that automatically identifies continuous-time transfer functions from Bode plots is presented. The identification is carried out in two stages. In the first one, the model order and 'good' guesses of the poles and zeros are obtained; in the second stage, estimates are refined by means of a modified Newton-Raphson algorithm. Because poles and zeros estimation only requires the magnitude curve, transport delays, if any, can be easily estimated by means of additional information supplied by the phase curve. The major and novel contribution of the proposed method resides in its first stage, where qualitative notions currently 'hidden' in the intuition of the designer are explicitly represented, yielding a simple optimization procedure that is not impaired by the presence of saddle points or local minima, and that converges very fast to the vicinity of the true solution. A detailed example is also provided to illustrate the value of the method.