利用多估计器输出间差异的目标跟踪算法及其仿真

Using the Difference Between Multiple Estimators' Outputs to Track Maneuvering Targets and Its Simulations

机动目标跟踪的难点在于如何在目标的各种机动情况下都有良好的跟踪性能。本文对目标跟踪的核心问题进行了研究,指出了当前统计模型中的系统方差自适应算法、利用新息的方法和模型切换方法的局限性,并提出了利用多个估计器输出之间的差异来选取合适输出的方法。仿真结果表明,该方法能够从多个估计器输出中选择出合适的输出,对目标的各种机动情况有较好的适应能力,计算量适中。

The key to track maneuvering targets is how to keep good performance while targets maneuver in various ways. The core of tracking maneuvering targets is studied in this paper. The limitations of current statistical method, innovation method, and changing model method are pointed out. A new method of tracking maneuvering targets is proposed. It uses the difference between multi-estimator抯 outputs to determine which is the best for output. The results of simulation prove that it can keep good performance while targets maneuver in various ways, and needs only moderate compute load.