Abstract: Aiming at the Modifiable Areal Unit Problem(MAUP) produced by grid clustering algorithm inevitably,an improved wavecluster algorithm based on double grid correction algorithm was presented.In this algorithm,the original grid clustering was performed when k value was obtained in the maximum number of dense grid;the corrected grid clustering was performed when k value was obtained in uniform distribution,thus the best clustering results were obtained with correction algorithm.An example of the rotor fault diagnosis based on double grid corrected wavecluster algorithm showed that the accuracy of double grid wave cluster algorithm was improved,which reduced the effect of grid partition and grid density on the clustering quality.The parallel processing of double grid clustering which could ease the conflict between even grid division and uneven distribution of data object provided a solution for MAUP.

Key words: wavecluster, modifiable areal unit problem, corrected algorithm, rotor, fault diagnosis

摘要： 针对网格聚类算法不可避免产生的网格量化问题,提出一种基于双网格校正小波聚类算法。该算法以最大密集网格规则下获得的网格划分k值进行原始网格小波聚类,以均匀分布假设规则下获得的网格划分k值进行校正网格小波聚类,通过校正算法得到最佳聚类结果。通过基于双网格校正小波聚类的转子故障诊断实例表明：双网格小波聚类降低了网格划分和网格密度阈值对聚类质量的影响,提高了精度;双网格聚类的并行校正处理缓和了一种尺寸下网格均匀划分与数据对象非均匀分布之间的矛盾,为网格量化问题提供了一种解决思路。

关键词: 小波聚类, 网格量化问题, 校正算法, 转子, 故障诊断