| 基于海气耦合模式的均权重粒子滤波与集合卡尔曼滤波比较研究 |
| |
| 引用本文： | 李科,苑福利,刘厂.基于海气耦合模式的均权重粒子滤波与集合卡尔曼滤波比较研究[J].海洋通报,2021,40(4). |
| |
| 作者姓名： | 李科 苑福利 刘厂 |
| |
| 作者单位： | 海军研究院,天津300061;哈尔滨工程大学智能科学与工程学院,黑龙江哈尔滨150001 |
| |
| 摘 要： | 在大气和海洋环境研究中,粒子滤波(PF)由于在非线性数据同化方面突出的优势,逐渐成为研究热点。最近改进的均权重粒子滤波(EWPF)为粒子滤波的进一步发展指明了新方向。集合卡尔曼滤波方法 (EAKF)作为当前主要应用的数据同化方法,使用高斯假设和线性假设来解决非线性问题,然而对均权重粒子滤波方法和卡尔曼滤波方法在非线性模式下的同化结果和特点还缺少系统详细的比较研究。本文在非线性耦合气候模式下,比较研究两种同化方法,采用均方根误差(RMSE)作为评价比较标准。实验结果表明,在非线性低频观测耦合模式中EWPF结果均优于EAKF。同时根据RMSE的结果得出,EWPF的同化结果更接近观察结果,而EAKF的同化结果更接近模式真值。
|
| 关 键 词： | 数据同化 海气耦合模式 粒子滤波 均权重粒子滤波 集合卡尔曼滤波 |
| 收稿时间： | 2020/12/17 0:00:00 |
| 修稿时间： | 2021/3/18 0:00:00 |
Comparison of Particle Filter and EAKF based on the ocean-atmosphere coupled model |
| |
| LI Ke,YUAN Fuli,LIU Chang.Comparison of Particle Filter and EAKF based on the ocean-atmosphere coupled model[J].Marine Science Bulletin,2021,40(4). |
| |
| Authors: | LI Ke YUAN Fuli LIU Chang |
| |
| Institution: | Naval Research Academy, Tianjin 300061, China;Collage of Intelligent Systems Science and Engineering, Harbin Engineering University, Harbin 150001, China |
| |
| Abstract: | Due to its advantage on nonlinearity, Particle Filter (PF) is rising as a promising advanced data assimilation method in the atmospheric and oceanic studies. The recently proposed, fully nonlinear equivalent weights particle filter (EWPF) represents the trend in particle filter development. The ensemble adjustment Kalman filter (EAKF) is also a famous algorithm which uses the Gaussian and linear assumptions to solve the nonlinear problems. However, comparing the EWPF and EAKF assimilation quality and their features are still incomplete and unsystematic. This paper systematically compares the performance of EWPF with traditional EAKF on the simple coupled climate model. The ensemble means and root-mean-square errors (RMSE) of model states are calculated as the metrics of the comparison. Generally, experiments show that the EWPF results are better than the EAKF in the two models. Base on the RMSE, the EWPF final assimilation is more consistent with the observations and the EAKF ensembles distribution is closer to the truth. |
| |
| Keywords: | data assimilation ocean-atmosphere coupled model particle filter EWPF EAKF |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
| 点击此处可从《海洋通报》浏览原始摘要信息 |
| 点击此处可从《海洋通报》下载免费的PDF全文 |
|
