Event-Triggered Particle Filter With One-Step Randomly Delayed Measurement

In this paper, we develop an event-triggered remote state estimator for nonlinear state-space models under networkinduced one-step randomly delayed measurements. Adopting event-triggering strategies reduces the transmission burden between the sensor and the estimator while maintaining the estimation accuracy. The developed method employs a particle filter to approximate the posterior distribution using particles and weights. In the non-triggering case, we use the constrained Bayesian estimation to compute the integrals associated with the posterior distribution. We evaluate the performance of the proposed algorithm using a simulated aircraft tracking problem. The results show that the proposed algorithm provides a comparable estimation accuracy to a particle filter without event triggering.