Abstract: Formal Concept Analysis（FCA），as a useful tool to deal with data，has recently received wide attention on the research areas.Because of the limits of two-value formal contexts，fuzzy set has been introduced into FCA.The primary problem coming with this is the enormous number of fuzzy concepts.The conceptual reduction of fuzzy dual concept lattices is discussed.Three methods to construct variable threshold dual concept lattices are given，and the relations of them are discussed.The results show that the number of concepts in variable threshold dual concept lattices is less than that in fuzzy dual concept lattices，and the important concepts are preserved.

Key words: Galois connection, fuzzy formal context, fuzzy dual concept lattice, variable threshold dual concept lattice

摘要： 形式概念分析作为进行数据分析的有效工具，吸引了世界各地研究者的关注。由于二值形式背景的局限性，人们将模糊集引入形式概念分析。但是模糊形式背景生成的模糊概念数量庞大。讨论了基于模糊形式背景的模糊对偶概念格的概念约简，构造出了模糊对偶概念格的三种变精度概念格，讨论了它们的关系。结果表明每一种变精度概念格都保留了模糊对偶概念格中相对重要的概念。变精度对偶概念格的概念数量远远少于模糊对偶概念格的概念数。

关键词: Galois连接, 模糊形式背景, 模糊对偶概念格, 变精度对偶概念格