In this paper we consider the tracking problem for a given action for a linear SISO plant, which model is presented in the input-output form. The problem is posed in the presence of noisy measurements and restrictions on the state variables and control action. The information about the derivatives of the given action is absent. The block synthesis procedure with nonlinear sigmoidal functions in feedbacks is developed for the solution of this problem. Whereas the estimation of state variables and measure-ment filtering are performed by using Kalman-Bucy filter.