Lattice Kalman Filters

In this paper, a new filter in the nonlinear Kalman filtering framework is proposed. The referred to as lattice (LKF) and based on class of quasi-Monte Carlo (QMC) methods known rules. proposed LKF method uses Korobov type rule deterministically generate sample points that are randomly shifted Cranley-Patterson shift order approximate multi-dimensional integrals Gaussian context. mathematical formulation well its error bound propagation discussed. To evaluate efficiency LKF, it applied aerospace system compared with four other well-known presented literature. Simulation results demonstrate significantly fewer sampling yielding lower computational burden than another variant QMC while maintaining estimation accuracy. Furthermore, provides asymptotically similar unscented (UKF) but less complexity, which an important consideration real applications.