Enhanced Hierarchical and Sequential Covariance Intersection Fusion
Zhongyao Hu, Bo Chen, Wen‐An Zhang, Li Yu
Abstract
Covariance intersection (CI) fusion is one of the most popular methods for combining estimates when the correlations among local estimation errors are unknown. Considering practical communication constraints, CI fusion tends to be performed in hierarchical and sequential forms, i.e., hierarchical CI (HCI) fusion and sequential CI (SCI) fusion. However, existing HCI and SCI fusion are sensitive to some uncertainties, i.e., the hierarchy structure and the fusion order, which make their fusion performances unreliable. To solve this problem, this article proposes hierarchy-structure-independent HCI fusion and fusion-order-independent SCI fusion by analogy with batch CI fusion, which can avoid possible negative effects caused by the aforementioned uncertainties. Finally, two simulations verify the effectiveness and advantages of the proposed methods.