带有信息反馈的最优异步递推航迹融合算法

现有的异步航迹融合算法大都采用无全局信息反馈的设计框架，并且忽略各异步传感器局部预测航迹误差间的相关性，加之考虑上述相关性的最优整体航迹融合算法的实时性难以被保证，从而导致局部传感器估计精度低、稳定性差、全局估计的最优性丧失和差的应用性等不足。为此，本文在引入全局信息反馈的基础上，同时考虑各局部预测航迹误差相关性，在融合估计误差协方差矩阵迹最小意义下，建立一种带有全局反馈机制的最优异步递推航迹融合算法。文中不仅对各种算法的次优或最优原理及性能进行了详细的分析与讨论，理论证明和仿真结果也表明了该文算法的有效性和优越性。

Optimal Asynchronous Recursive Track Fusion with Global Feedback

Asynchronous fusion is common in processing multisensor data fusion problems and the asynchronous track fusion which adopts distributed fusion architecture has an extensive application among asynchronous fusion. In current asynchronous track fusion algorithms, the global feedback is mostly not established from fusion center to local sensors, so the local estimate can not be improved by the global estimate. Another defect among these algorithms is the correlation between the local prediction track estimate errors is ignored; thereby the accuracy of global estimate is reduced. In order to improve the estimate precision of the fusion system, the correlation is considered and the global feedback is introduced in this paper, accordingly a novel optimal asynchronous track fusion algorithm with feedback is presented. Compared with the current algorithm without global feedback and correlation, the proposed method can improve the estimate performance of the fusion system. The algorithm analysis and simulations both show the advantages of the novel algorithm.