Federated ensemble Kalman filter in no reset mode design

Abstract

The main contribution of this paper is to design a more accurate optimal/suboptimal fault tolerant state estimator. Federated filters compose of a set of local filters and a master filter, the local filters work in parallel and their solutions are periodically fused by the master filter yielding a global solution. Federated ensemble Kalman filter no reset configuration is developed for multi-sensor data fusion. Ensemble Kalman filter(ENKF) estimation is widely used, where the models are of extremely high order and nonlinear, the initial states are highly uncertain, and a large number of measurements are available. ENKF is used as local filters in federated filter no reset mode design. Fault detection and isolation (FDI) algorithms is applied to local filter's outputs. Faulty local filters are isolated and not fused by master filter to get a fault tolerant filter. Simulation results demonstrate the validity of the proposed filter formation.