【摘要】 根据测量误差和状态估计误差,采用数论法、随机点集包络法和数据拟合方法计算测量状态的跟踪门形状和大小。跟踪门是多传感器多目标关联算法的首要技术,实质是对多目标进行分布范围划分,减小目标的无效检测率。现有跟踪门的门限算法以各个测量分量或分量函数具有正态分布特性为前提,加之规范的形状,具有特定的局限性。数论法根据误差源的概率分布采用在Weyl意义偏差较小的均匀分布点集来模拟误差源的统计分布。该方法适合于三维空间跟踪门计算,计算量小,过程简单,不仅可得到跟踪门形状,而且可对跟踪门的大小进行计算,达到减少交叉跟踪门总数、交叉跟踪门内目标总数的目的。更多还原

【Abstract】 According to measurement covariance and state estimation covariance,this paper give out tracking gate shape and size through introduction of Number Theory Method(NTM),random dot assemble envelope method and polynomial fit method.In essence,in order to decrease the inefficacy detection probability,tracking gate separate the distribution of targets into several parts which is the chief issue of multi-sensor and multi-target tracking problem.The existing tracking gate algorithm is deficient for the hypothesis of normal distribution and uniform shape.NTM introduces uniform distribution dot assembly in the sense of Weyl to simulate the distribution of variance source under specific distribution probability density.NTM could give out not only the shape but also the size of tracking gate for three-dimensional space with comparably minimize latency to decrease the number of intersectional tracking gate and targets in the intersectional tracking gate.更多还原