Nonstationary spectral peak estimation by Monte Carlo filter

The aim of research is to estimate multiple peaks of power spectrum that ore varying with time. A model to estimate the time-varying peaks has been proposed by the author. The model is written in a state space representation composed by a system model and an observation model. The system model denotes smooth change of a state vector that consists of pairs of peak frequency and bandwidth. The observation model is autoregressive model with time-varying coefficients that are nonlinearly parametrized by the state vector. The nonlinear parametrization is based on a fact that the pairs of frequency and bandwidth are roots of characteristic equation of the autoregressive model. Estimating the state vector by giving the observations results the estimation of frequency and bandwidth pairs of time-varying pourer spectrum. In store estimation, properties of nonlinear and non-Gaussian should be treated because of the nonlinear formulation of the model. As a method of state estimation, we have employed an approximation of non-Gaussian distribution by its realizations, called Monte Carlo filter. Through numerical examples, estimation precision of peak frequency has been checked by comparing with a conventional model.