Gaussian framework for nonlinear state estimation with stochastic event-trigger and packet losses

Abstract

Nonlinear state estimation with stochastic event-trigger and packet losses is studied in this paper. To handle the nonlinearity and the uncertainty of the available information, the recursive probability density functions (PDFs) are characterized as Gaussian. Then, an event-trigger and packet losses induced Gaussian filter (EPGF) and its Gaussian smoother (EPGS) are derived to develop a new Gaussian framework. This developed framework is an extension of the standard Gaussian one and suitable for both linear and nonlinear systems. The key to its implementation is calculating a series of integrals with the Gaussian weight form, which can be approximated by various numerical techniques. In addition, according to the rule of three-degree spherical-radial cubature, two specific filtering and smoothing algorithms of the proposed framework are presented. Finally, a numerical simulation illustrates the effectiveness of the developed scheme.