A Hybrid–Robust Method for Estimating the Values of Linear Functionals under Essential Uncertainty and the Nonlinear Factor

For secondary data processing tasks, a hybrid-robust method is developed to estimate various numerical characteristics (linear functionals) optimally under parametric uncertainty factors related to the information process, regular and irregular interferences, and the correlation function of measurement noise (including the case of nonlinear parameters). The method is fully or partially invariant with respect to these factors and yields the desired estimates without expanding the vector of estimated parameters or employing linearization procedures. A comparison with known estimation methods is presented, and the random and methodological errors, as well as the achieved computational effect, are analyzed. An illustrative example is provided.