On the Wiener Prediction Problem and the Problem of the Optimized, Nosiefree Automatic Controller

摘要

A solution of the Wiener prediction problem in the sampled, non-stationary case has been presented by Kalman. (1) The method used is founded on statistical orthogonal projections. Kalman and Koepcke (2) (3) have also presented a solution of the problem of minimizing a generali-zed loss-function defined for a sampled, noisefree automatic controller. Here dynamic programming is used. The solutions of the two problems have structural resemblances and the problems are called the duals of each other by Kalman, who finds the close resemblance very surprising.