This work presents a mean-square error lower bound for state estimation of nonlinear stochastic systems under given differentiable state constraints. Its recursive formulation permits incorporation of random process and measurement errors and is shown to be a generalization of the known lower bound for unconstrained problems. The bound is evaluated for the example of locating a ground vehicle from noisy measurements of its horizontal position and velocity incorporating a roadmap.
%0 Journal Article
%1 schmitt17constrainedCrlb
%A Schmitt, Lorenz
%A Fichter, Walter
%D 2017
%J IEEE Signal Processing Letters
%K 2017 ifr journal myown
%N 12
%R 10.1109/LSP.2017.2764540
%T Cramér-Rao Lower Bound for State-Constrained Nonlinear Filtering
%V 24
%X This work presents a mean-square error lower bound for state estimation of nonlinear stochastic systems under given differentiable state constraints. Its recursive formulation permits incorporation of random process and measurement errors and is shown to be a generalization of the known lower bound for unconstrained problems. The bound is evaluated for the example of locating a ground vehicle from noisy measurements of its horizontal position and velocity incorporating a roadmap.
@article{schmitt17constrainedCrlb,
abstract = {This work presents a mean-square error lower bound for state estimation of nonlinear stochastic systems under given differentiable state constraints. Its recursive formulation permits incorporation of random process and measurement errors and is shown to be a generalization of the known lower bound for unconstrained problems. The bound is evaluated for the example of locating a ground vehicle from noisy measurements of its horizontal position and velocity incorporating a roadmap.},
added-at = {2017-10-24T10:01:39.000+0200},
author = {Schmitt, Lorenz and Fichter, Walter},
biburl = {https://puma.ub.uni-stuttgart.de/bibtex/2fec019dd9eb1973eabfb348a7fde312e/lorenzschmitt},
doi = {10.1109/LSP.2017.2764540},
interhash = {b80beb7d5e945fe193c0710af58b312e},
intrahash = {fec019dd9eb1973eabfb348a7fde312e},
journal = {IEEE Signal Processing Letters},
keywords = {2017 ifr journal myown},
month = {December},
number = 12,
timestamp = {2017-11-05T08:15:40.000+0100},
title = {Cramér-Rao Lower Bound for State-Constrained Nonlinear Filtering},
volume = 24,
year = 2017
}