An Adaptive Scheme Based on Distributed Kalman Filter for Sensor Network
Aiming at reducing the kinematic model error and uncertain measurement noise in sensor network,an adaptive scheme for distributed Kalman filter (DKF) is proposed in this paper.An adaptive factor is firstly applied to the covariance matrix of the predicted state vector to make the covariance estimation agree with its theoretical one,thus eliminating the effect of kinematic model error.Based on the innovation covariance of measurement noise and an adjustable fading factor,the other adaptive strategy is further developed for the updating of the covariance matrix of the measurement noise,which contributes to reduce the effect of the uncertain measurement noise.As demonstrated in simulation results,the adaptive distributed Kalman filter (ADKF) for sensor network perforrns better than the plain DKF in the case of suffering the kinematic model error and uncertain measurement noise.
Adaptive distributed Kalman filter sensor network consensus filters
Zhang Kezhi
shanghai aircraft design and research institute
国际会议
沈阳
英文
1381-1384
2012-07-27（万方平台首次上网日期，不代表论文的发表时间）