Fusion Estimation Of SICI States For Networked Systems With Packet Loss

In recent years,the traditional sensor systems have developed rapidly in the direction of networking.There are two magic weapons in the multi-sensor networked systems: network and distributed structure,and under the support of these superiorities,the research of multi-sensor networked systems has attracted much attention of academic circles.Different from the traditional sensor systems,the network can expand the application of the sensor systems,and the distributed structure can improve the flexibility and reliability of the sensor systems.However,the transmission of network data cannot be completed 100%.Due to the influence of aging of equipment performance,physical circuit failure,software BUG or network virus,etc.,network failures such as packet dropouts or random delays may occur during data transmission,and these problems often appear randomly.For this reason,how to eliminate these random network problems is the primary task of studying multi-sensor networked systems.On the other hand,as a multi-sensor fusion architecture,distributed multi-sensor networks require reliable and accurate multi-sensor information fusion algorithms to ensure the accuracy of the fusion results.Therefore,the multi-sensor fusion estimation problem of networked systems has become a research hot spot both at home and abroad.In this paper,focus on the network problem with packet dropouts in the networked sensor systems,the sequential inverse covariance intersection(SICI)fusion estimation algorithm was proposed to solve the problems of networked systems with packet dropouts.The main research contents of this paper are as follows:Firstly,for the multi-sensor networked system with packet dropouts and linearly correlated white noises,the local time-varying Kalman predictor and filter were proposed by using state expansion and fictious noises technology.Then the local steady-state Kalman filter was proposed by the Lyapunov equation method.Based on above work,the inverse covariance intersection(ICI)fusion steady-state Kalman filter was designed,and combine it with the sequential fusion method,the sequential inverse covariance intersection(SICI)fusion steady-state Kalman filter was derived.Secondly,for the multi-sensor networked system with packet dropouts and multiplicative noises,the local time-varying and steady-state Kalman estimators,ICI fusion steady-state Kalman estimators and SICI fusion steady-state Kalman estimators were designed.Finally,for the multi-sensor networked system with packet dropouts and random delay,the local time-varying and steady-state Kalman estimators,ICI fusion steady-state Kalman estimators,and SICI fusion steady-state Kalman estimators were designed.