I. V. Semushin, J. V. Tsyganova, K. V. Zakharov, “Robust filter algorithms — survey and new results for ship navigation and conning systems”, Informatsionnye Tekhnologii i Vychslitel'nye Sistemy, 2013, Issue 4,Pages <nobr>90

APPLIED ASPECTS OF COMPUTER SCIENCE

Robust filter algorithms — survey and new results for ship navigation and conning systems

I. V. Semushin^a, J. V. Tsyganova^a, K. V. Zakharov^b

^a Ulyanovsk State University
^b FRPC OJSC ‘RPA ‘Mars’

Abstract: Discrete filtering is considered with emphasis on computational aspects. A short survey of numerically stable filter implementations is given founded on the basic mathematical ideas: positive definite (covariance or information) matrix factorization, vector measurement scalarization, matrix triangularization and array-matrix orthogonalization. Using them, LD-algorithm of Extended Kalman Filtering is proposed as applied to the non-linear problem of target motion analysis. Application of the algorithm to ship navigation and conning (including collision avoidance) is discussed.

Keywords: computer data engineering, algorithms of least squares method, discrete filtering, filter divergence, numerical stability, extended Kalman filter, specific statistical applications and models, digital simulation and system modeling.