一种改进的蝴蝶算法优化粒子滤波算法

(西安科技大学 通信与信息工程学院,陕西 西安 710054)

粒子滤波; 粒子贫化; 状态估计; 蝴蝶算法; 滤波精度

An improved butterfly algorithm optimizing particle filter algorithm
ZHANG Wei-hu,GUO Ming-xiang,HE Yuan-kai,SUN Xiao-ting,ZHU Dai-xian

(College of Communication and Information Engineering,Xi'an University of Science and Technology,Xi'an 710054,China)

particle filtering; particle depletion; state estimation; butterfly algorithm; filtering accuracy

DOI: 10.13800/j.cnki.xakjdxxb.2019.0117

备注

传统的粒子滤波算法在重采样期间丢弃小重量粒子,因此重要性权重落在极少数粒子上。这会导致采样粒子贫化、粒子多样性缺失以及需要大量粒子才能进行比较准确的状态估计等问题,针对这些问题,提出了一种改进的蝶式算法优化粒子滤波算法。首先,将最新时刻观测信息引入蝴蝶香味公式中,以提高滤波精度; 其次,引入吸引半径参数来控制蝴蝶种群寻优的搜索范围,降低算法的复杂度,进而提高算法的实时性; 最后,将改进的蝴蝶种群位置更新公式用于优化迭代更新。实验结果表明,与经典粒子滤波器和现有蝶形优化算法相比,改进算法具有更低的均方误差和运行时间。并且在粒子数较少的情况下,可以实现更准确的状态估计,并改善传统滤波器的粒子耗尽现象,保证了粒子多样性。

Traditional particle filtering algorithms discard small weight particles during resampling,so importance weights fall on very few particles,leading to the problem of sampled particle depletion,lack of particle diversity,and the need for a large number of particles for more accurate state estimation.In this paper,an improved butterfly algorithm is proposed to optimize the particle filter algorithm.Firstly,the latest moment observation information is introduced into the butterfly flavor formula to improve the filtering accuracy; Secondly,the attraction radius parameter is introduced to control the search range of butterfly population optimization,which reduces the complexity of the algorithm and improves the real-time performance of the algorithm.Finally,the improved butterfly population location update formula is used to optimize iterative updates.The experimental results show that compared with the classical particle filter and the existing butterfly optimization algorithm,the improved algorithm has a lower mean square error and running time.And in the case of a small number of particles,more accurate state estimation can be achieved,and the particle depletion phenomenon of the conventional filter can be improved,and the particle diversity is ensured.