| 摘要: |
| 针对长江口地区某测波站实测连续波面过程资料中出现的异常值问题，探讨了判别资料中异常数据的几种方法。由编制的计算机程序比较表明：用肖维勒准则，采用分次判别异常值的方法可对原始数据直接进行检索和判别异常值，在实际资料的应用中效果良好。 |
| 关键词: 异常值,判别方法,肖维勒准则,海洋资料 |
| DOI： |
| 分类号:P717 |
| 基金项目: |
|
| Methods used to analyse and distinguish abnormal data in oceanic information |
|
Wang Hongchuan, Zuo Qihua, Pan Junning
|
|
Nanjing Hydraulic Research Institute
|
| Abstract: |
| Aimed at the unusual data appearing in the information of the wave surface in one of measuring stations of Yangtze river estuary were discussed some kinds of methods to distinguish the abnormal data in oceanic information. Compared these methods by computer and concluded that according to Chauvenet criterion the method could examine and distinguish the abnormal data in initial informations by iteration, and has good effect in practical application. |
| Key words: abnormal data, methods of judgment, Chauvenet criterion, oceanic information, |
