| 基于本征正交分解的湍流去噪 |
| |
| 引用本文： | 卢垚,刘欢,吴加学.基于本征正交分解的湍流去噪[J].海洋学报,2022,44(9):132-144. |
| |
| 作者姓名： | 卢垚 刘欢 吴加学 |
| |
| 作者单位： | 1.中山大学 海洋科学学院, 广东 珠海 519082 |
| |
| 基金项目： | 国家自然科学基金联合基金（U1901209）；国家自然科学基金面上项目（41676082）；南方海洋科学与工程广东省实验室（珠海）创新团队建设项目（311021004） |
| |
| 摘 要： | 在近岸河口湍流研究中，实测的湍流资料往往容易受到噪声的影响，导致湍流特征量的估算出现偏差。本征正交分解是一种将流场在能量上进行分解再重构的方法，基于该方法，结合数值实验和野外实测资料，对包含噪声的湍流数据进行了去噪处理。结果表明：（1）本征正交分解能有效去除湍流中的噪声。在进行信号重构时，应将保留能量的百分比与湍流能量占比保持一致。去噪的效果与噪声占比有关，噪声占比越高，去噪效果越明显。（2）在实测资料中，憩流时刻的噪声占比要显著大于非憩流时刻，水平方向的噪声占比要大于垂直方向。经过本征正交分解去噪后，各湍流特征量的估算更加合理。
|
| 关 键 词： | 本征正交分解 湍流 去噪 能谱 湍流特征量 |
| 收稿时间： | 2021-09-09 |
Turbulence denoising based on proper orthogonal decomposition |
| |
| Institution: | 1.School of Marine Science, Sun Yat-sen University, Zhuhai 519082, China2.Southern Marine Science and Engineering Guangdong Laboratory (Zhuhai), Zhuhai 519080, China |
| |
| Abstract: | In the studies of turbulence in coastal and estuary, the turbulence data from field observation is often influenced by noise, which leads to deviate the estimation of turbulence properties. Proper orthogonal decomposition is a method which decompose and reconstruct the flow structure in energy filed. Based on this method, the turbulence data containing noise is denoised combined with numerical experiments and field observation data. The results show that: (1) Proper orthogonal decomposition can remove the noise in turbulence effectively. The percentage of retained energy in reconstruction should be consistent with the percentage of turbulence energy in the raw signal. The results of denoising are related to the percentage of noise, the higher the percentage of noise, the effect of denoising is more obvious. (2) In the field observation, the percentage of noise in the slack tide is obviously higher than the noise which in the non-slack tide, while the percentage of noise in the horizontal direction is higher than that in the vertical direction. The estimation of each turbulence properties is more reasonable after using the proper orthogonal decomposition to denoise. |
| |
| Keywords: | |
|
| 点击此处可从《海洋学报》浏览原始摘要信息 |
| 点击此处可从《海洋学报》下载免费的PDF全文 |
|
