序统计滤波估计检测海洋测深异常数据

Order Statistics Filtering for Detecting Outliers in Depth Data Along a Sounding Line

根据海洋水深数据处理的要求,构建了海洋水深数据的序统计滤波模型,并根据水深异常数据的特点,提出了离差法判定水深异常数据,在遵循“舍深取浅”这一基本海洋数据处理原则基础上,设计了海洋水深异常数据检测的序统计滤波方案。实例分析表明,该滤波方案能够有效地检测并剔除零水深、负水深、孤立突跳水深,保留连续(个数多于(含)3个的)异常水深,在序统计滤波异常数据检测的基础上剔除粗差,能大幅提高粗差检测的效率。

According to the special demands in processing depth data, the models based on order statistics filter were presented here to detect outliers in depths along a sounding line. However,there is a key problem of how to distinguish the outliers from good depths;it was solved by the size of the differences between a would-be-filtered point and its neighbor points. And then a complex method based on order statistics filter to detect outliers in depths was given out to meet the various needs in processing sounding data. At last, the method has been tested using observed data. The results show that the method could remove gross errors, zero and minus depths, and preserve consecutive （more than 3 ） false echoes. And it could help the surveyors find and remove outliers effectively.