Linear adaptive controller based on smoothing noisy data algorithm

Spline approximation of signals together with convolution technique of modelling are used to develop a linear adaptive controller for continuous time systems. The presented approach allows to construct the adaptive controller of LQG type by a simple modification of the discrete LQG algorithm. The resulting controller is believed to cover a rather board class of control problem, such as control of system with complex models, distributed parameter systems and can be used in multirate control.