Generalized work functional in regularization of the estimation of the state of a dynamical system

It is shown that estimation of the state of a dynamical system from measure ments obtained in the course of a single session is an ill-posed problem. I t is proved that the Krasovskii generalized work functional regularizes the initial problem. An example illustrating the process of regularization tha t uses an algorithm with predictive model is presented, and an empirical de pendence between the prediction interval and the optimal values of the regu larization parameter is obtained for the example.