| 改进的粒子滤波算法在船用组合导航中的应用 |
| |
| 引用本文： | 江健,李伟峰,姚健,史国友.改进的粒子滤波算法在船用组合导航中的应用[J].上海海事大学学报,2018,39(2):17-21. |
| |
| 作者姓名： | 江健 李伟峰 姚健 史国友 |
| |
| 作者单位： | 大连海事大学航海学院;大连海事大学航海安全保障重点实验室 |
| |
| 基金项目： | 国家自然科学基金（51579025）；辽宁省自然科学基金（20170540090） |
| |
| 摘 要： | 将典型的粒子滤波算法应用于非线性INS/GPS船用组合导航系统中时,不能解决重要性函数与似然函数的不匹配问题,且当传感器具有较大的漂移误差时,无法根据当前的观测值对粒子状态进行修正。为解决这些问题,基于典型的粒子滤波算法提出一种改进的粒子滤波(improved particle filter,IPF)算法。根据当前的观测值估计对应的状态误差,并在确定粒子的权值和重采样前对预测的粒子的误差进行修正。仿真结果表明:IPF算法能够有效解决上述问题,保持很好的滤波精度。
|
| 关 键 词： | 惯性导航 组合导航 误差修正 粒子滤波算法 |
| 收稿时间： | 2017/11/26 0:00:00 |
| 修稿时间： | 2018/1/9 0:00:00 |
Application of an improved particle filter algorithm in integrated navigation for ships |
| |
| Institution: | Navigation College of Dalian Maritime University,Navigation College of Dalian Maritime University,Navigation College of Dalian Maritime University and Navigation College of Dalian Maritime University |
| |
| Abstract: | When the typical particle filter algorithm is applied to the nonlinear integrated navigation system INS/GPS for ships, it can not solve the mismatch between the importance function and the likelihood function, and particle states can not be corrected according to current observation values when sensors have large drift errors. In order to solve the problems, an improved particle filter (IPF) algorithm is proposed based on the typical particle filter algorithm. The corresponding state errors are estimated according to current observation values, and the errors of predicted particles are corrected before determining the weights and resampling the particles. The simulation results show that the IPF algorithm can effectively solve the above problems and maintain good filtering accuracy. |
| |
| Keywords: | inertial navigation integrated navigation error correction particle filter algorithm |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《上海海事大学学报》浏览原始摘要信息 |
| 点击此处可从《上海海事大学学报》下载免费的PDF全文 |
