论文部分内容阅读
针对拟蒙特卡洛粒子滤波(QMC-PF)算法计算量大,难以满足雷达目标跟踪实时性需要的问题,提出了一种适用于雷达机动目标跟踪的新型拟蒙特卡洛粒子滤波算法(NQMC-PF).该算法利用QMC方法生成权重较大粒子的低差异性的子代粒子来替换低权重粒子,保证了样本的质量和多样性,同时利用广义回归神经网络(GRNN)计算子代粒子的权重,提高了滤波的精度和速度.实验结果表明,该算法的计算精度高于标准拟蒙特卡洛粒子滤波算法,同时运算时间短,实时性好,能够应用于对雷达目标的跟踪上.
Aiming at the problem that the quasi-Monte Carlo particle filter (QMC-PF) algorithm is computationally intensive and can not meet the real-time requirement of radar target tracking, a new quasi-Monte Carlo particle filter algorithm (NQMC- PF). This algorithm replaces the low-weight particles with the QMC method to generate the low-variance particles, which guarantees the quality and diversity of the samples. At the same time, the GRNN Weight and improve the accuracy and speed of the filtering.Experimental results show that the proposed algorithm has higher computational accuracy than the standard Monte Carlo particle filter algorithm and has the advantages of short computation time and good real-time performance, and can be applied to the tracking of radar targets.