Accession Number : AD0701427

Title :   A SENSITIVITY ALGORITHM FOR THE KALMAN FILTER AND PREDICTOR,

Corporate Author : TEXAS UNIV AUSTIN ELECTRONICS RESEARCH CENTER

Personal Author(s) : Lainiotis,Demetrios G. ; Railey,Malcolm R.

Report Date : FEB 1970

Pagination or Media Count : 30

Abstract : In this investigation sensitivity matrix algorithms for the continuous-time Kalman filter and predictor are derived. Nonlinear matrix differential equations of the Riccati type are derived. The solutions of these equations represent the increase in the error covariance matrix of a suboptimal estimator over the optimal estimator. These solutions permit the sensitivity analysis to be performed a priori on linear estimation problems. The objective of this investigation was to provide the designer a technique to simplify the sensitivity analysis of the Kalman filter and predictor. (Author)

Descriptors :   (*INFORMATION THEORY, MATHEMATICAL PREDICTION), SENSITIVITY, ALGORITHMS, WHITE NOISE, NONLINEAR DIFFERENTIAL EQUATIONS

Subject Categories : Cybernetics

Distribution Statement : APPROVED FOR PUBLIC RELEASE