A sequential method of nonlinear parameter estimation for random processes

A method is proposed for confidence parameter estimation in random processes which are nonlinearly represented in the process equations (stochastic differential or difference). The results are applied to estimation of slow drift in extremal control in the presence of noise.