论文部分内容阅读
为了提高雷达对机动目标的跟踪精度,通过融合拟蒙特卡罗思想,提出了一种适用于非线性非高斯系统的拟蒙特卡罗粒子滤波交互式多模型算法。该算法利用拟蒙特卡罗采样,克服传统算法采样粒子间隙过大、粒子层叠问题,增加交互式多模型对机动目标跟踪时的有效粒子数;通过区间估计理论,解决拟蒙特卡罗支撑区间难以计算问题,并结合核密度估计重采样,保证采样粒子的低等差异性。仿真结果表明:与交互式多模粒子滤波算法相比,改进算法可在保证滤波实时性的同时,提高跟踪精度。
In order to improve the tracking accuracy of the maneuvering target, a quasi-Monte Carlo particle filter interactive multi-model algorithm for non-linear non-Gaussian systems is proposed by fusing the quasi-Monte Carlo method. The algorithm uses quasi-Monte Carlo sampling to overcome the problem of oversize and particle stacking in the traditional algorithm and increases the number of effective particles when the interactive multi-model is used to track the maneuvering target. The interval estimation theory solves the problem that quasi-Monte Carlo support interval is difficult Calculate the problem, and combined with nuclear density estimation re-sampling, to ensure low sampling particle diversity. The simulation results show that compared with the interactive multi-mode particle filter algorithm, the improved algorithm can improve the tracking accuracy while ensuring the real-time filtering performance.