摘要
为了提高双网格校正小波聚类算法的效率,提出了基于散列函数的双网格校正小波聚类算法,该算法应用散列表去消除量化数据空间中的空单元,降低数据空间算法的复杂度。先量化特征空间;再构造散列函数形成散列表,将量化后的特征值存储到散列表中;在散列表上并行对原始网格和校正网格进行小波变换,在特征空间的不同层次上寻找连通单元;利用校正网格产生的聚类结果去校正原始网格产生的聚类结果,得到最终聚类结果。将此方法应用到航空发动机转子故障诊断中,实验证明此算法在保证了转子故障诊断精度的基础上,提高了效率。
A double-grid correction wavelet clustering algorithm based on hash function was proposed to improve the efficiency of the double-grid correction wavelet clustering one. A hash table was used to eliminate empty cells in a quantized data space and reduce the complexity of the data space algorithm. Firstly,the eigen-space was quantized.Secondly,the hash function was constructed to form a hash table to store quantized eigenvalues in the hash table. Then,the wavelet transform was conducted for the original grid and the corrected one on the hash table simultaneously to find connected elements in different layers of the eigen-space. Finally,the clustering results generated by the corrected grid were used to correct those generated by the original one,and the final clustering results were obtained. This method was applied in rotor fault diagnosis of an aircraft engine. The test results showed that the proposed algorithm improves the rotor fault diagnosis’ s efficiency based on ensuring its accuracy.
作者
刘晓波
韩子东
邵伟芹
左红艳
LIU Xiaobo;HAN Zidong;SHAO Weiqin;ZUO Hongyan(School of Aeronautical Manufacturing Engineering,Nanchang Hangkong University,Nanchang 330063,China;School of Mechanical and Electrical Engineering,Nanjing University of aeronautics and Astronautics,Nanjing 210016,China)
出处
《振动与冲击》
EI
CSCD
北大核心
2018年第21期267-272,280,共7页
Journal of Vibration and Shock
基金
国家自然科学基金(51365040)
航空科学基金(2013ZD56009)
江西省自然科学基金(20151BAB206060)
关键词
小波聚类
散列函数
双网格校正
故障诊断
wavelet clustering
hash function
double-grid correction
fault diagnosis
