A Multi-Feature Fusion Method for Fault Diagnosis with Applications in Tail Paddle Lever Bearing

Aiming at the problems such as difficult monitoring and inaccurate state evaluation of helicopter tail rotor rod bearing. In this paper, temperature and vibration sensors are used to collect the temperature and vibration signals of bearings. Then the collected signal is constructed by means of noise reduction, feature extraction and feature fusion. And the characteristic data set is used as the input of fault diagnosis model after state index evaluation, health index generation and dimensionality reduction. Last, the improved gravity search algorithm (GGSA) is used to optimize the parameters of support vector machine (SVM) diagnosis model to form GGSA-SVM model for bearing fault diagnosis. In the simulation test, compared with the previous method, the number of iterations of the model is reduced by 21 times, the fitness is reduced to 0.4%, and the accuracy of bearing fault diagnosis is increased by 4.4% to 97.9<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">%</sup>. The experimental results show that the model can be used for bearing fault diagnosis better.