State Estimation of The Underwater Moving Target Based on Multi-Sensor Information Fusion
The Kalman filter has been widely used in state estimation of moving targets. Furthermore, the wellknown conventional Kalman filter requires an accurate system model and exact prior information. However, the nonlineariry, dynamic and randomicity in the underwater environment result in uncertainty and incontinuity of the observation information and unknown bias of the system model, which may seriously degrade the performance of the Kalman filter or even cause the filter to diverge. Therefore, a novel filtering algorithm based on multisensor information fusion estimation theory and dynamic bayesian network inference is presented, which is based on the idea of fusing firstly and then filtering. Firstly, the multi-sensor system fuses the sub-systems measurement information to obtain more accurate initial measurement information and covariance information, and then smooth missing data and fuzzy data by fusing the the obtained system state predictive information and all the measurement information of the subsystems to obtain the accurate state estimation of underwater moving target. Finally, the simulation results show that the proposed method can efficiently estimate underwater target state without prior noise information, and can evidently improve the state estimation precision of underwater moving target by adjusting the weight factor despite noise-related.
state estimation underwater moving target multi-sensor information fusion
Tang Zheng Sun Chao Liu Zong-wei Meng Di
Institute of Acoustical Engineering Northwestern Polytechnical University Xian China Northwestern Polytechnical University Xian China
国际会议
秦皇岛
英文
416-419
2010-11-05（万方平台首次上网日期，不代表论文的发表时间）