摘要
通过对已有粗差探测成果的研究,本文按照基于的学科不同将其大致分为4类:数理统计粗差探测方法,基于多元统计分析粗差探测方法,线性代数粗差探测方法,人工智能粗差探测方法。然后从这4个方面辐射出已有的关于粗差发现、定位、定值等相关方法的研究,并且针对国内学者对用人工智能方法实现粗差探测未进行成熟研究的现状,提出可行策略并进行实现从而为人工智能和测量数据处理2个学科的结合开拓新的研究方向。
By the research of the existed achievement on gross error detection, this paper divided them into four categories in accordance with the different disciplines on which they based： the statistics-based method, the multivariate statistical-based method, the linear algebra-based method, the artificial intelligence-based method. And then the paper elaborated the research about finding and lo- cating gross error and valuing as well which are radiated from the four areas. Furthermore, for the status that the domestic scholars have not been achieved maturity research about using artificial intelligence to detect gross errors, the possible strategies were proposed and realized, in order to explore new research directions for the combination of the artificial intelligence and the data processing.
出处
《测绘科学》
CSCD
北大核心
2012年第5期14-16,共3页
Science of Surveying and Mapping
基金
支持增量更新的分布式异构空间数据无缝集成技术研究与软件开发(2007AA12Z204)
关键词
粗差探测
数理统计
多元统计
线性代数
人工智能
gross error detection
statistics
multivariate statistical
linear algebra
artificial intelligence
