GROUP OBJECT STRUCTURE AND STATE （组对象结构和状态）.pdf

GROUP OBJECT STRUCTURE AND STATE ESTIMATION IN THE PRESENCE OF MEASUREMENT ORIGIN UNCERTAINTY Lyudmila Mihaylova and Amadou Gning Lancaster University, Dept. of Communication Systems, InfoLab21, Lancaster LA1 4WA, UK Email: mila.mihaylova@lancaster.ac.uk, e.gning@lancaster.ac.uk ABSTRACT cluding leader-follower models [7]. However, estimating the dynamic evolution of the group structure has not been widely This paper proposes a technique for motion and group structure es- timation of moving targets based on evolving graph networks [1] studied in the literature, although there are similarities with in the presence of measurement origin uncertainty. The proposed methods used in evolving network models [11, 12]. method, through an evolving graph model, allows to jointly esti- In [1], the group structure is modeled as evolving undi- mate the group target and the group structure with the uncertainty. rected random graphs with measurements which origin is ex- The performance of the algorithm is evaluated and results with real actly known. Then, an evolution model is defined for the ground moving target indicator data are presented. group structure and the efficiency of the approach is showed through a scenario with simulated data. The main contribu- Index Terms— Evolving graphs, random graphs, group tion of the present paper is the proposed solution for group target tracking, nonlinear estimation, Monte Carlo methods, object structure and state estimation with measurement uncer- data associatio