基于统计决策规则提出自适应采样数粒子滤波算法, 在定义综合性能风险函数的基础, 推导出粒子数与滤波误差方差之间的关系式, 使得在跟踪过程中, 可以根据目标的机动情况在线调节粒子数, 以使跟踪性能

达到最优。在Matlab仿真平台下进行了闪烁噪声下的机动目标跟踪实验, 结果表明, 自适应采样数粒子滤波算法是一种有效的机动目标跟踪方法, 跟踪性能较基本粒子滤波算法提高了3.17倍。-Based on statistical decision rules of the number of adaptive sampling particle filter algorithm, in defining the basis for comprehensive performance risk function derived particle number and the filtering error variance of the relationship between the type, makes the tracking process, it can be depending on the target-line adjustment of the motor case particle number in order to achieve optimal tracking performance. In the Matlab simulation platform, carried out the flicker noise under the maneuvering target tracking experiment results show that the number of particle filter algorithm for adaptive sampling is an effective method of maneuvering target tracking, tracking performance than the elementary particle filter algorithm increases 3.17-fold.