Identifiability Analysis of Noise Covariances for LTI Stochastic Systems with Unknown Inputs
Abstract
Most existing works on optimal filtering of linear time-invariant (LTI) stochastic systems with arbitrary unknown inputs assume perfect knowledge of the covariances of the noises in the filter design. This is impractical and raises the question of whether and under what conditions one can identify the process and measurement noise covariances (denoted as and , respectively) of systems with unknown inputs. This paper considers the identifiability of / using the correlation-based measurement difference approach. More specifically, we establish (i) necessary conditions under which and can be uniquely jointly identified; (ii) necessary and sufficient conditions under which can be uniquely identified, when is known; (iii) necessary conditions under which can be uniquely identified, when is known. It will also be shown that for achieving the results mentioned above, the measurement difference approach requires some decoupling conditions for constructing a stationary time series, which are proved to be sufficient for the well-known strong detectability requirements established by Hautus.
Keywords
Cite
@article{arxiv.2209.07104,
title = {Identifiability Analysis of Noise Covariances for LTI Stochastic Systems with Unknown Inputs},
author = {He Kong and Salah Sukkarieh and Travis J. Arnold and Tianshi Chen and Biqiang Mu and Wei Xing Zheng},
journal= {arXiv preprint arXiv:2209.07104},
year = {2022}
}
Comments
formally accepted to and going to appear in IEEE Transactions on Automatic Control. arXiv admin note: substantial text overlap with arXiv:2202.04963