Optimization of recurrence algorithms for estimating the parameters of linear plants

A technique for optimal, in terms of asymptotic rate of convergence, choice of parameters for algorithms of estimating the type of stochastic approximation is suggested.