Crack Fault Classification for Planetary Gearbox Based on Feature Selection Technique and K-means Clustering Method

Abstract

During the condition monitoring of a planetary gearbox, features are extracted from raw data for a fault diagnosis. However, different features have different sensitivity for identifying different fault types, and thus, the selection of a sensitive feature subset from an entire feature set and retaining as much of the class discriminatory information as possible has a directly effect on the accuracy of the classification results. In this paper, an improved hybrid feature selection technique (IHFST) that combines a distance evaluation technique (DET), Pearson's correlation analysis, and an ad hoc technique is proposed. In IHFST, a temporary feature subset without irrelevant features is first selected according to the distance evaluation criterion of DET, and the Pearson's correlation analysis and ad hoc technique are then employed to find and remove redundant features in the temporary feature subset, respectively, and hence, a sensitive feature subset without irrelevant or redundant features is selected from the entire feature set. Further, the k-means clustering method is applied to classify the different kinds of health conditions. The effectiveness of the proposed method was validated through several experiments carried out on a planetary gearbox with incipient cracks seeded in the tooth root of the sun gear, planet gear, and ring gear. The results show that the proposed method can successfully distinguish the different health conditions of a planetary gearbox, and achieves a better classification performance than other methods. This study proposes a sensitive feature subset selection method that achieves an obvious improvement in terms of the accuracy of the fault classification.

A Filter Feature Selection Algorithm Based on Mutual Information for Intrusion Detection

“…Intrusion detection based on networks is an important step of cyber security [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15]. By analyzing large amounts of network data, network-based intrusion detection can effectively mitigate security threats [16][17][18][19][20][21][22][23][24][25][26][27][28][29][30][31][32][33][34][35]. Therefore, data processing plays a vital role in intrusion detection.…”

“…According to different evaluation functions, the filter methods are divided into five categories: distance, information (or uncertainty), dependence, consistency and the classifier error rate [15]. In recent years, many feature selection algorithms have been proposed [16][17][18][19][20][21][22][23][24][25][26][27][28][29][30][31][32][33][34][35]. When feature selection is applied properly, it can significantly improve classification processing time and performance.…”

“…Features selected by different feature selection algorithms using the denial of service (DOS) classifier. ,4,19,15,16,18,17,31,14,28,23,20,26,11,27,40,3,29,1,42,13,32,38,30,39,5,41 MIFS (β = 0.7 )[19] 317, 2, 19, 15, 16, 18, 17, 14, 23, 20, 31, 26, 28, 11, 27, 40, 29, 3, 1, 42, 39, 30, 4, 5, 32, 13, 41, 36, 38, 35, 34 , 13, 4, 19, 15, 16, 17, 18, 14, 28, 20, 23, 31, 29, 11, 26, 27, 40, 42, 3, 38, 1 …”

“…Features selected by different feature selection algorithms using the R2L classifier. ,10,14,15,16,19,21,22,20,4,17,12,18 MIFS (β = 0.7 )[19] 208,9,10,14,15,16,19,21,22,20,12,18,4,17,26,27,23,31,30,11 …”

Theoretical Guidance on Evacuation Decisions after a Big Nuclear Accident under the Assumption That Evacuation Is Desirable

“…Ray and Turi used K-means cluster analysis to classify data between groups and within groups and applied it to color image segmentation [41]. Wang and Shao used K-means to classify the fault feature information of a planetary gearbox in order to divide the planetary gearbox's health status into different levels [42]. Desarbo et al [43] first applied K-means to weight calculations.…”