Information filtering and its application to relative navigation

Abstract

PurposeThe purpose of this paper is to estimate the relative states between the leader and wingman based on vision‐based relative navigation system using extended information filtering (EIF).Design/methodology/approachFor a typical leader‐wingman formation case, the relative navigation equations are introduced. Vision‐based navigation system which consists of an optical sensor and a series of specific light sources is used to capture the line‐of‐sight measurement between the two unmanned aerial vehicles (UAVs). Owing to the limitations on the field of view of the optical sensor, not every specific light source would be visible. And the spatial relative position of the two vehicles could also contributes to the diminution of visibility since some of the light sources are likely to be shielded by the frame and wing. Therefore, the EIF can be applied to the vision‐based relative navigation while every specific light source is regarded and processed as an individual information source. It is demonstrated that the information of visual source could be easily extracted by the simple update equation of information filtering.FindingsThe EIF could be used in vision‐based relative navigation system to give an accurate estimation of relative position, velocity and attitude without increasing the amount of calculation or decreasing the estimation accuracy compared to conventional Kalman filtering.Originality/valueThe EIF is first introduced to the vision‐based relative navigation in order to provide relative state between UAVs during formation flight.