A Probabilistic Model-adaptive Approach for Tracking of Motion with Heightened Uncertainty
- PDF / 820,047 Bytes
- 12 Pages / 594.77 x 793.026 pts Page_size
- 82 Downloads / 150 Views
ISSN:1598-6446 eISSN:2005-4092 http://www.springer.com/12555
A Probabilistic Model-adaptive Approach for Tracking of Motion with Heightened Uncertainty J. Josiah Steckenrider* and Tomonari Furukawa Abstract: This paper presents an approach for state tracking in scenarios where motion is highly uncertain. The proposed approach improves on traditional Kalman filters by integrating model parametric uncertainty in deriving state covariances for prediction at each time step. A model correction stage then continuously adjusts the mean and variance of state matrix elements based on the observation-corrected state, compensating for an initially inadequate system model. The symbiotic relationship between state tracking and motion model correction is leveraged to perform both tasks simultaneously in-the-loop. In a representative dynamic example, simulated experiments were performed and analyzed statistically for varying combinations of sensor and model uncertainty. For low model variance, traditional Kalman filters generally perform estimation better due to over-confidence with regards to model parameters. However, the proposed approach increasingly outperforms both traditional and adaptive Kalman filters in estimation when model and input uncertainty is appreciable. The motion model updating approach formulated here tends to improve parameter estimates over the course of state tracking, thus validating the symbiotic process. The robotics applications of this simultaneous estimation and modeling framework extend from target state tracking to self-state estimation, while broader signal processing applications can be readily extracted. Keywords: Kalman filter, motion model updating, online filtering, recursive Bayesian estimation, robotic target tracking, state estimation.
1.
INTRODUCTION
1.1. Background In probabilistic state estimation by an autonomous robot, from drone target tracking and pursuit to pose estimation of a mechatronic manipulator, existing filtration methods are well-suited to handle uncertainty. Some techniques are even capable of overcoming non-Gaussian and nonlinear processes. Probabilistically equipped autonomous robots often employ a form of recursive Bayesian estimation (RBE) to construct belief and estimate states with inherent uncertainties. State estimation falls apart when parameters informing a predictive model are inaccurately estimated [1]. The work presented here seeks to remedy this issue by incorporating model uncertainty in RBE prediction, as well as introducing a specially formulated model-updating step. This framework, termed Simultaneous Estimation and Modeling (SEAM), is developed in this paper for systems experiencing Gaussian or near-Gaussian state uncertainty. 1.2. Related work RBE consists of the recursive iteration of three stages
[2] which, while known by various terms in different communities, are referred to here as prediction, observation, and correction. Various simplifying approximations lead to different versions of RBE, of which the Kalman filter (KF) [3] and its variants
Data Loading...
