Filtering in Stochastic Dynamical Systems When There is Anomalous Noise in the Observation Channel. Ⅱ Systems with a Continuous-Discrete Observation Channel

摘要

The problem of synthesizing a root-mean-square-optimal filter that is unbiased in the class of linear filters is solved in the case of a set of timewise continuous and discrete filters when anomalous noise is present in the discrete observation channel. The sensitivity of the filter to the intensity matrix of the anomalous noise, the optimality of the procedure for eliminating anomalous measurements, and the accuracy of estimation in the case of different sets of anomalous observation channels are investigated. The case of systems with standby of the observation channels is examined.