可处理任意步延迟OOSMs的最优更新算法

发布时间：2018-07-09 21:46

本文选题：目标跟踪系统 + 无序量测　；参考：《电光与控制》2017年05期

【摘要】：在目标跟踪系统中,因通信延迟会出现传感器量测无序到达融合中心的现象,从而产生无序量测问题。而在系统实际工作过程中,面临的常常是多个无序量测问题。针对此问题,对其常见情形进行分类,在前向预测框架内提出了Dl算法。该算法对无序量测融合周期内各时刻状态估计与协方差矩阵进行更新,可处理任意步延迟无序量测。此外,针对其中单个无序量测,推导出融入等价量测的信息滤波器,无需求状态转移矩阵的逆,且不依赖过程噪声离散化模型。仿真结果验证了算法的精确性与有效性。
[Abstract]:In the target tracking system, because of the communication delay, the sensor measurement will arrive at the fusion center out of order, which leads to the disorder measurement problem. In the actual working process of the system, it is often faced with multiple disordered measurement problems. In order to solve this problem, this paper classifies the common cases and puts forward the D1 algorithm in the framework of forward prediction. The algorithm updates the state estimation and covariance matrix at each time during the fusion period of unordered measurements, and can deal with any step delay disorder measurement. In addition, for a single unordered measurement, an information filter incorporating equivalent measurements is derived, the inverse of the demandless state transition matrix is derived, and the discretization model is independent of process noise. Simulation results verify the accuracy and effectiveness of the algorithm.
【作者单位】：军械工程学院信息工程系;
【分类号】：TN713

【相似文献】