Abstract: Planetary gearboxes are highly susceptible to failure as mechanical transmission equipment. In order to reduce the number of sensor arrangements in planetary gearbox fault diagnosis such that reduce the cost, planetary gearbox measurement point optimization method based on the multi-dimensional ensemble empirical mode decomposition (MEEMD) information entropy combined with correlation analysis is proposed. Firstly, the vibration test signals for the five operating conditions are decomposed using MEEMD. Secondly, the correlation coefficients between the decomposed components and the raw data are used to filter out the IMF components containing the main fault information, and their information entropy features are extracted to construct the sample feature vector. Finally, the information entropy eigenvectors of different measurement points of the same operating condition and different operating conditions of the same measurement point are controlled for correlation analysis. The relatively redundant measurement points are analyzed and eliminated to achieve the goal of measurement point optimization.