Kalman filter algorithms in complex navigation system of the airplane

Practical realisation of a Kalman filter (KF) in the complex navigation systems (CNS) assumes that statistical characteristics of meters´ errors are known. In real conditions the prior information is of approximated nature. Sometimes it´s necessary to lower the requirements to the onboard digital computer (ODC) of the airplane: the volume of the basic memory and calculation time. In this case the model of KF input signal is simplified. Finally, filter output estimations become nonoptimum estimations, as the matrix transmission factor K(t) of KF is calculated irrespective of the acting measurements. Therefore at KF designing in CNS it is necessary to evaluate the decrease of complex accuracy, which can take place in real-life environment. Thus, it is necessary to evaluate error stability of KF of CNS. Analysed in this paper are the problems of KF design (2003). The KF error stability in the CNS considered is estimated. Characteristics of standard ODC for realization of KF optimum and suboptimum algorithms are analysed. It is shown that realisation of optimum and suboptimum integration algorithms obtained in considered CNS of an airplane does not complicate modern ODC.