摘要
针对标准粒子滤波算法存在的缺陷,本文引入了两种改进的方法,引入最新的量测信息,改进粒子滤波的建议分布。EKPF通过引入扩展卡尔曼算法改进粒子分布,UPF引入无味变换改进粒子的分布,并对其进行了仿真对比分析。实验结果表明,UPF算法优于扩展卡尔曼粒子滤波算法与标准粒子滤波算法。
For the defects of the standard particle filter algorithm, two improved algorithm are proposed, which introduced the latest measurement and improved particle filter proposal distribution. EKPF through the introduction of extended Kalman al- gorithms to improve the particle distribution, UPF introduces the unscented transformation to improve the distribution of the par- ticles, and a simulation of comparative analysis is given. The experimental results show that the UPF algorithm is better than the extended Kalman particle filter algorithm and the standard particle filter algorithm.
出处
《自动化与仪器仪表》
2013年第1期10-13,共4页
Automation & Instrumentation
基金
国家自然基金项目(61263031)
甘肃省自然科学基金项目(1014ZSB064)
甘肃省财政厅项目(0914ZTB148)
