信息重构融合深度学习的泵机组故障诊断方法OA

Aiming at the problems of noise interference and low diagnostic accuracy of pump unit in actual operation,a fault diagnosis method of pump unit based on information reconstruction and deep learning was proposed.Firstly,the Red-billed blue magpie optimizer(RBMO)was used to optimize Variational mode decomposition(VMD),to decompose the vibration signal into a series of subsequences.Secondly,the vibration signal was reconstructed using Permutation Entropy(PE)and filtering algorithm.Finally,the generated image was input into a neural network based on the EfficientNet network to output the diagnostic results,and the Motif difference field(MDF)was used to convert the reconstructed signal into a two-dimensional image.This method was verified by experiments.Experimental results show that the proposed method can accurately classify faults,and has better noise resistance and higher accuracy than other methods under the condition of strong noise interference.