论文部分内容阅读
针对粒子滤波中存在的重要性密度函数难以选取和可能出现粒子退化的问题,提出了一种改进渐消中心差分粒子滤波算法.该算法从研究粒子滤波的建议分布出发,充分利用最新观测信息,通过引入改进的渐消因子实时调整增益阵,得到一种优化的建议分布函数,能够有效缓解粒子退化现象.同时,在粒子采样中,研究采用了多样化的采样方式减轻滤波中粒子枯竭的影响,进一步提高了滤波精度.将该算法应用到单站无源定位系统中进行仿真运算,仿真结果表明,在不同的观测精度环境下,该算法使得系统具有更好的自适应性和滤波精度.
Aiming at the problem that the importance density function existed in the particle filter is difficult to select and the particle degeneration may occur, an improved fading center differential particle filter algorithm is proposed. This algorithm starts from the study of particle filter recommendation distribution, makes full use of the latest observation information, By introducing an improved fading factor to adjust the gain matrix in real time, an optimized proposed distribution function is obtained, which can effectively mitigate the particle degeneration phenomenon.At the same time, a variety of sampling methods are used to reduce the influence of particle exhaustion in the particle sampling , Which further improves the filtering accuracy.The algorithm is applied to single-station passive positioning system to simulate the simulation results show that the algorithm has better self-adaptability and filtering accuracy under different observation accuracy environment.