Scalable event-triggered distributed extended Kalman filter for nonlinear systems subject to randomly delayed and lost measurements

In this paper, a distributed extended Kalman filter (EKF) is developed for a class of nonlinear systems, whose outputs are measured by multiple sensors which send data using an event triggered mechanism through a communication network subject to loss and latency. Random transmission delay and multiple dropouts are modelled by a Bernoulli random sequence. The filter gains are determined in each sensor node such that an upper bound on the cross covariance of the estimation error is minimized; so, less computational burden is required, even in the networks with the large number of nodes. To be specific, the scalability is the main feature of the proposed scheme. The boundedness of the filtering error is proved under some conditions. Finally, comparative simulation results are presented to illustrate the effectiveness and the applicability of the suggested filter. (C) 2020 Elsevier Inc. All rights reserved.