Estimation of Quasipolynomials in Noise: Theoretical, Algorithmic and Implementation Aspects
Vytautas Slivinskas, Virginija Šimonytė · Birkhäuser Boston eBooks · 1997
The problem of parametric estimation of the sum of real-valued quasipolynomials in white Gaussian noise is considered. We discuss theoretical and algorithmic aspects of this problem. A real quasipolynomial is defined as a polynomial with real coefficients multiplied by an exponential function with a real exponent. The polynomial coefficients vary in time. Their varying is described by a sinusoidal function. The main focus is given to estimation of the damping factors and frequencies of the quasipolynomial model from noisy data. In this paper, we also give a short description of software developed by the authors which deals with the estimation problem mentioned above. Numerical results are presented.