Prediction and compensation of errors of inertial navigation systems

An original approach is proposed to the solution of prediction problems, the main peculiarity of which is in the creation of the relatively compact algorithms of long term prediction, based on the combination of the system state variables or the laws of their application, which remain constant at the prediction interval. The suggested prediction procedure is not based on the complication of the initial model of the system or process, but on the more accurate description of the variation laws of their parameters. The resulting concept is applied to the solution of the problem of extrapolation of heavy plane inertial navigation system (INS) errors after the external navigation sensors are switched off.<
>