Abstract: As an extension of the multi-granularity rough set, a rough analysis model of weighted multi-granularity based on limited dominance relation is proposed. Firstly, the limitation of the rough set based on dominance relation and multi-granularity rough set is analyzed and the limited dominance relation is introduced to improve the weighted multi-granularity rough set for high-dimensional or distributed systems with incomplete and reference attribution, making the preference relations of properties and the difference among the granularities fully considered. The rough approximations of know-ledge are acquired and the relationship with the optimistic and pessimistic model is discussed. Then, relevant theorems and properties are proved. Finally, the experimental results show the feasibility and effectiveness of the method.

Key words: incomplete information, multi-attribute decision system, limited dominance relation, weighted multi-granularity rough set

摘要： 作为多粒度粗糙集的推广，提出了限制优势关系下的加权多粒度粗糙分析方法。分析了优势关系粗糙集和多粒度粗糙集的局限性，引入限制优势关系对加权多粒度粗糙集进行改进，充分考虑到属性的偏好关系和粒度的重要性差别，使其适用于不完备的高维或分布式有序信息系统；由此得出粗糙近似，讨论了其与限制优势关系下的乐观、悲观多粒度粗糙集的关系，并对相关性质和定理进行证明；通过实例验证了该方法的有效性和实用性。

关键词: 不完备信息, 多属性决策系统, 限制优势关系, 加权多粒度粗糙集