Abstract:
A version of the duality principle in Kalman's filter theory is formulated whereby the problem of determining the optimal statistical estimate of the phase coordinate vector for a dynamic system with incomplete knowledge in the presence of noise is equivalent to the problem of determining the optimal estimate of the phase coordinate vector with incomplete data without noise and with indirect constraint on the time of esimate computation. The duality principle is used to consider the adjustment algorithm in an adaptive system with dynamic response stabilization optimal in terms of minimal mean square errors of adjustment with specified time of the transient process in the adjustment loop.