Filtering in Stochastic Systems: Analysis for the case of continuous observations with memory of arbitrary multiplicity

International Conference on Information Technologies in Business and Industry 2016

计算机科学;经济学;工业技术

Rozhkova, O.V.^1 ; Demin, N.S.^2 ; Li, J.^3 ; Moldovanova, E.A.^1 ; Natalinova, N.M.^1 ; Genkel, V.A.^1

Tomsk Polytechnic University, 30, Lenina ave., Tomsk

634050, Russia^1

Tomsk State Technical University, 36, Lenina ave., Tomsk

634050, Russia^2

Tsinghua University, 30 Shuangqing Rd, Haidian, Beijing

100084, China^3

关键词: Anomalous component;    Channel with memory;    Continuous observation;    Continuous-time;    Equivalence to;    Mean square;    Noise intensities;    Observation vectors;   

Others  :  https://iopscience.iop.org/article/10.1088/1742-6596/803/1/012129/pdf
DOI  :  10.1088/1742-6596/803/1/012129

来源: IOP

【 摘 要 】

We consider stochastic systems with continuous time over observations with memory in the presence of an anomalous noise. The paper is devoted to analysis of some properties of an optimal unbiased in mean-square sense filter. In the case of anomalous noises action in the observation channel with memory, we have proved insensitivity of the filter to inaccurate knowledge of the matrix of anomalous noise intensity and its equivalence to a truncated filter constructed only over non-anomalous components of an observation vector.