Cubature quadrature filter for one-step randomly delayed measurements

Abhinoy Kumar Singh, Shovan Bhaumik, Kundan Kumar

Abstract: The paper proposes an extension of the conventional cubature quadrature Kalman filter (CQKF), in order to enable it for dealing with randomly delayed measurement problems for nonlinear systems. The proposed extension is bounded with an assumption that the maximum delay does not exceed unit sampling time which is alternatively called as one step delayed measurements in this paper. The proposed method which uses third-degree spherical cubature rule and arbitrary order Gauss-Laguerre quadrature rule to solve the intractable integrals is more accurate compared to its existing cubature Kalman filter (CKF) based counterpart. The efficacy of the proposed algorithm has been demonstrated for a maneuvering target tracking problem