摘要： 针对专家系统知识难以获取的问题,提出了利用粗糙集理论对飞行数据模式特征进行提取的方法,分别采用自组织映射方法对属性进行离散化、利用遗传算法进行属性约简,最后在处理结果中引入专家的领域先验知识。实验结果表明,这种方法是可行的。

关键词: 粗集, 飞行数据, 特征提取

Abstract: This thesis proposes a method of extracting pattern attributes of flight data based on rough set theory, which is aimed to solve thedifficulty in constructing knowledge base using the expert system. The method uses self-organizing map to get the discrete attributes first, thensimplifies the information through genetic algorithm, and finally uses prior knowledge of domain experts to process the simplified information.Results of experiment prove the feasibility of this method.

中图分类号: 

• TP182