基于确定性重采样的粒子滤波算法

Particle Filter Algorithm Based on Deterministic Resampling

复杂背景下的运动目标跟踪往往要面对非线性非高斯问题,粒子滤波算法在非线性非高斯模型中的良好处理能力,使其得到广泛的应用.引入重采样方法(SIR)解决粒子退化问题的同时导致了样本枯竭.针对上述问题,文中提出了一种融合基本重采样方法和确定性重采样方法的新方法,能有效保持粒子的多样性.通过仿真实验表明,该方法能有效提高粒子滤波算法的准确性.

Because of its good performance in the complex background of the non-linear and non Gaussian model,the particle filter algorithm has ben used widely. Resampling（SIR） leads to sampledepletion when it is used to solve degeneracy of particles. In order to solve the problem above, this paper presents a new method which integrates the basic resampling method and the deterministic resampling method. The simulation results show that the new method can not only effectively maintain the diversity of particles but also improve the correctness of the particle filter algorithm.