论文部分内容阅读
在非线性非高斯动态系统中,粒子滤波已成为解决系统参数估计和状态滤波的主流方法。然而,粒子退化是粒子滤波中不可避免的现象,粒子重采样是解决方法之一。本文针对粒子退化现象,在扩展卡尔曼滤波器的基础上研究了一种基于支持向量机粒子滤波算法,算法实现中扩展卡尔曼粒子滤波器结合支持向量机对当前时刻的重要性采样,再对粒子样本进行重采样。该算法能有效地利用量测值的最新信息,状态估计误差较小,同时避免了粒子匮乏。理论分析和仿真结果表明,新算法在双模噪声非线性系统估计的精度优于标准粒子滤波算法与扩展卡尔曼粒子滤波算法。
In non-linear non-Gaussian dynamic systems, particle filtering has become the mainstream method to solve system parameter estimation and state filtering. However, particle degeneration is an unavoidable phenomenon in particle filtering. Particle resampling is one of the solutions. In this paper, based on the extended Kalman filter, a particle swarm optimization algorithm based on Support Vector Machine (SVM) is studied in this paper. The importance of the extended Kalman filter combined with SVM to the current time is sampled. Particle samples are resampled. The algorithm can effectively use the latest information of the measured value, the state estimation error is small, while avoiding the particle scarcity. The theoretical analysis and simulation results show that the accuracy of the new algorithm is better than the standard particle filter and the extended Kalman filter in the estimation of the dual-mode noise nonlinear system.