388 0

Improved target tracking using Gaussian mixture measurements for long-range radar and passive sonar systems

Title
Improved target tracking using Gaussian mixture measurements for long-range radar and passive sonar systems
Author
Qian Zhang
Advisor(s)
Taek Lyul Song
Issue Date
2017-08
Publisher
한양대학교
Degree
Doctor
Abstract
Target tracking with nonlinear measurements is a common and challenging problem in radar and sonar systems. The possibility and accuracy of target tracking in such systems are determined by adopting what kind of technique to address the nonlinearity of measurements. The tracking filter using Gaussian mixture measurements (GMM) is a novel and efficient technique to handle the nonlinear measurements, and is based on converted measurements and Gaussian mixture model. In this thesis, improved target tracking using GMM for long-range radar and passive sonar systems is proposed. In long-range radar tracking systems, the measurement uncertainty region has a thin and curved shape in Cartesian space due to the fact that the measurement is accurate in range but inaccurate in angle. Such a shape reflects grievous measurement nonlinearity, which can lead to inconsistency in tracking performance and significant tracking errors in traditional nonlinear filters, such as extended Kalman filter (EKF), unscented Kalman filter (UKF), and particle filter (PF). A modified version of GMM is proposed to efficiently handle severe nonlinearity in 2-D and 3-D long-range radar tracking. In passive sonar systems, only bearings-only measurements are utilized to track targets. Such measurements also suffer from severe nonlinearity. For single-target tracking, a novel iterated GMM filter is proposed to represent measurement likelihood function with Gaussian mixtures more precisely. The proposed approach recalculates range interval using updated track components, and the likelihood function is remodelled using Gaussian mixtures with new range interval. For multi-target tracking, an improved nonlinear filter, GMM-Probability Hypothesis Density (GMM-PHD), is proposed by combining GMM with the well-known PHD filter. The GMM-PHD can provide high tracking accuracies for bearings-only multi-target tracking.
URI
http://hdl.handle.net/20.500.11754/33584http://hanyang.dcollection.net/common/orgView/200000430980
Appears in Collections:
GRADUATE SCHOOL[S](대학원) > ELECTRONIC SYSTEMS ENGINEERING(전자시스템공학과) > Theses (Master)
Files in This Item:
There are no files associated with this item.
Export
RIS (EndNote)
XLS (Excel)
XML


qrcode

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.

BROWSE