Abstract:
Construction of algorithm of extended Kalman filter for a nonlinear continuous - discrete problem of an estimation of a state vector of system is described. It is considered the problem of the adaptive filter where the estimations of the covariance matrices is carried out for noises of process and measurements with the purpose to increase an accuracy of a state vector's estimation. It is shown the schemes for several types of adaptive algorithms and results of mathematical modelling of their work for a rigid body with given inertia.