canonical decomposition of random process

Optimal forecast algorithm based on compatible linear filtration and extrapolation

This work considers the methods of optimal linear extrapolation of the flight path of the aircraft, which provide a minimum of the mean square of the forecast error with different amounts of a priori information.  The research is based on the canonical decomposition of a vectorial random process.  It is determined that the development of modern technologies entails increasing requirements for quality and accuracy of control.  However,  since the existing methods of linear extrapolation do not provide for the maximum accuracy of the forecast due to the inherent constraints on the random proc