Mean-shift-based robust distributed set-membership fusion filtering for sensor network systems with outliers

Abstract

Outliers can contaminate the communication and measurement processes of many sensor network systems, which may be induced by environmental disturbances, model uncertainties, sensor faults or errors, subnetwork faults or malicious cyberattacks. Once the distributed set-membership filter (DSMF) is used into such sensor network systems with outliers for distributed state estimation, the estimation performance can be seriously degraded in each sensor node. To address this problem, this article proposes a mean-shift-based trust set and zonotope extraction mechanism to modify the zonotopic DSMF toward building resilience and robustness against outliers. The proposed mechanism is capable of sifting out the outlier-contaminated local corrected zonotopes, and a sufficient condition for the full effectiveness of it is given and proved. Based on the proposed mechanism, the mean-shift-based outliers-robust zonotopic DSMF (MSRDSMF) is derived for estimating the state of a sensor network system in a distributed way. Simulation experiment results demonstrate the practical validity and superiority of the MSRDSMF in effectively suppressing the effects of outliers.