Parameter estimation for nonlinear stochastic model using generalized entropy optimization principle

A new type of parameter estimation method has been proposed for a class of nonlinear stochastic model with non-Gaussian disturbance The Parzen window method was first used to estimate the density function of the sampled data and then the generalized entropy optimization principle was used to estimate the unknown parameters. No matter what distribution the noise obeys to, Gaussian or non-Gaussian, unbiased parameter estimated values can be obtained. The simulation results show the effectiveness of the proposed approaches.