Nonlinear filtering algorithms for the parameters of the state vector of a
dynamic system described by stiff stochastic differential equations are syn
thesized via a new approach in which stiffness is avoided and, consequently
, the filter takes less time to generate the estimate of the state vector.