In this paper, the prediction-based distributedfiltering problem is discussed for a class of time-varying stochastic systems with communication delay and different types of noises over sensor networks. The communicat...
详细信息
ISBN:
(纸本)9789881563972
In this paper, the prediction-based distributedfiltering problem is discussed for a class of time-varying stochastic systems with communication delay and different types of noises over sensor networks. The communication delay is characterized when the state estimations are transmitted between adjacent sensor nodes. In order to compensate the effects induced by the communication delay, the prediction-based idea is employed and then the active compensation estimation is provided when designing the time-varying distributed filter. In particular, both the prediction-based state estimation and its own innovation measurements are utilized in terms of the concerned sensor networks under given topological structure. Subsequently, a locally minimum upper bound of the filtering error covariance is given by determining the filter gain at each time step. Finally, the validity and advantages of the presented prediction-based distributedfiltering method are illustrated by some simulations.
In this paper, the prediction-based distributedfiltering problem is discussed for a class of time-varying stochastic systems with communication delay and different types of noises over sensor networks. The communicat...
详细信息
In this paper, the prediction-based distributedfiltering problem is discussed for a class of time-varying stochastic systems with communication delay and different types of noises over sensor networks. The communication delay is characterized when the state estimations are transmitted between adjacent sensor nodes. In order to compensate the effects induced by the communication delay, the prediction-based idea is employed and then the active compensation estimation is provided when designing the time-varying distributed filter. In particular, both the prediction-based state estimation and its own innovation measurements are utilized in terms of the concerned sensor networks under given topological structure. Subsequently, a locally minimum upper bound of the filtering error covariance is given by determining the filter gain at each time step. Finally, the validity and advantages of the presented prediction-based distributedfiltering method are illustrated by some simulations.
暂无评论