Stochastic Event-Triggered Sequential Fusion Filtering for USV Cooperative Localization
Mengfei Niu, Guanghui Wen, Han Shen, Yuezu Lv, Guanrong Chen
Abstract
This article deals with the cooperative localization of maneuvering unmanned surface vessel (USV) based on multisensor fusion estimation, in which a sequential fusion filter is designed to estimate the real-time position of the USV. To avoid excessive communication consumption between sensors and the fusion filter, a stochastic event-triggered communication mechanism is adopted to ensure necessary measurements transmission. With the aid of the classical framework of sequential Bayesian filtering, an event-triggered sequential fusion filter is constructed by codesigning the stochastic event-triggered communication mechanism and the sequential filter, where a technique of unscented transformation with the sequential idea is used to resolve the intractable problem caused by nonlinear measurement models. Furthermore, a sufficient condition is established to ensure the boundedness of the fusion covariance. Finally, the effectiveness and superiority of the designed fusion filter is verified both by numerical simulation and practical experiment of a real USV tracking system.