Data Association for Simultaneous Localization and Mapping in Dynamic Environments using Multiple Model Approach
Item
-
Title
-
Data Association for Simultaneous Localization and Mapping in Dynamic Environments using Multiple Model Approach
-
Identifier
-
d_2009_2013:76d9480c8842:11461
-
identifier
-
11845
-
Creator
-
Wong, Rex H.,
-
Contributor
-
Jizhong Xiao | Mumtaz Kassir
-
Date
-
2012
-
Language
-
English
-
Publisher
-
City University of New York.
-
Subject
-
Electrical engineering | Robotics | data association | dynamic clutter | interactive multiple models | joint probability | optimization assignment | SLAM
-
Abstract
-
Simultaneous localization and mapping is the fundamental problem in mobile robotics. Data association is considered the most difficult part of this problem. The difficulties come from two major sources: the uncertainty of positions and measurements; and the dynamics of the environments. In addition, time-constraint and computational complexity further hinder the development of data association algorithms in real-time applications. In this work, we propose a feasible approach to handle the real-time data association problem in clutter and dynamic environment. The approach is three-folded. First, we characterize this problem as an optimization assignment problem with the constraint of one-to-one assignment in presence of clutter and interference. The mathematical models are based on Bayesian Joint Probability Data Association and using the linear programming algorithms to obtain the feasible assignment solutions. Second, we adopt the concept of multiple model system so that each tracking filter can adaptively estimate the target behavior more accurately when the mode of relative motion changes. Third, the incorrect decision on initialization may be revoked by a specific designed sliding window method and ranked assignment formulation. Because of its adaptability and cost-effectiveness, this approach can be applied for real-time SLAM applications. Simulation is performed to demonstrate the effectiveness of this method.
-
Type
-
dissertation
-
Source
-
2009_2013.csv
-
degree
-
Ph.D.
-
Program
-
Engineering
