Motion Planning for Active Data Association and Localization in Non-Gaussian Belief Spaces

2020 ◽  
pp. 288-303
Author(s):  
Saurav Agarwal ◽  
Amirhossein Tamjidi ◽  
Suman Chakravorty
Keyword(s):  
Motion Planning ◽  
Data Association ◽  
Active Data ◽  
Non Gaussian

Stochastic Optimal Motion Planning for the Attitude Kinematics of a Rigid Body With Non-Gaussian Uncertainties

2014 ◽  
Vol 137 (3) ◽  
Author(s):  
Taeyoung Lee
Keyword(s):  
Rigid Body ◽  
Motion Planning ◽  
Uncertainty Propagation ◽  
Planck Equation ◽  
Uncertain Environments ◽  
Stochastic Motion ◽  
Stochastic Optimal Control Problem ◽  
Optimal Motion Planning ◽  
Non Gaussian ◽  
Global Uncertainty

This paper investigates global uncertainty propagation and stochastic motion planning for the attitude kinematics of a rigid body. The Fokker–Planck equation on the special orthogonal group is numerically solved via noncommutative harmonic analysis to propagate a probability density function along flows of the attitude kinematics. Based on this, a stochastic optimal control problem is formulated to rotate a rigid body while avoiding obstacles within uncertain environments in an optimal fashion. The proposed intrinsic, geometric formulation does not require the common assumption that uncertainties are Gaussian or localized. It can be also applied to complex rotational maneuvers of a rigid body without singularities in a unified way. The desirable properties are illustrated by numerical examples.

scholarly journals Probabilistic Data Association-Feedback Particle Filter for Multiple Target Tracking Applications

2017 ◽  
Vol 140 (3) ◽  
Author(s):  
Tao Yang ◽  
Prashant G. Mehta
Keyword(s):  
Particle Filter ◽  
Target Tracking ◽  
Data Association ◽  
Multiple Target Tracking ◽  
Multiple Target ◽  
Probabilistic Data ◽  
Probabilistic Data Association ◽  
Particle Filter Algorithm ◽  
Gaussian Case ◽  
Non Gaussian

This paper is concerned with the problem of tracking single or multiple targets with multiple nontarget-specific observations (measurements). For such filtering problems with data association uncertainty, a novel feedback control-based particle filter algorithm is introduced. The algorithm is referred to as the probabilistic data association-feedback particle filter (PDA-FPF). The proposed filter is shown to represent a generalization—to the nonlinear non-Gaussian case—of the classical Kalman filter-based probabilistic data association filter (PDAF). One remarkable conclusion is that the proposed PDA-FPF algorithm retains the error-based feedback structure of the classical PDAF algorithm, even in the nonlinear non-Gaussian case. The theoretical results are illustrated with the aid of numerical examples motivated by multiple target tracking (MTT) applications.

Extending and evaluating the posture-based model of motion planning

2006 ◽  
Author(s):  
Jonathan Vaughan ◽  
Steven Jax ◽  
David A. Rosenbaum
Keyword(s):  
Motion Planning
Sign in / Sign up

Export Citation Format

Share Document