Model Modification and State Estimation for a Nolinear System

This paper considers a nonlinear system containing some unknown parameters. When random perturbations are introduced,the first correction terms of its solutions canbe proved to be Gaussian processes,and then the measured date can be used to construct alikelihood function and to estimate unknown deterministic or stochastic parameters.One example of state estimation in biochemical engineering is also given to illustrate the main idea.