Centrifugal pump bearing life prediction method and device based on VMD-GRU

The invention discloses a centrifugal pump bearing life prediction method and device based on VMD-GRU. The method comprises the following steps: acquiring a fault signal of a target bearing; the target bearing is arranged in the multi-stage centrifugal pump; performing variational mode decomposition on the fault signal to obtain a component signal of the fault signal; inputting the component signal into a prediction model to obtain the residual life of the target bearing output by the prediction model; the prediction model is obtained based on gated recurrent neural network training. The one-dimensional fault signal of the centrifugal pump bearing can be decomposed into the multi-dimensional fault component signal, and the residual life of the centrifugal pump bearing is predicted based on the decomposed fault component signal, so that the accuracy of predicting the life of the centrifugal pump bearing is improved. 本申请公开了一种基于VMD－GRU的离心泵轴承寿命预测方法、装置。该方法包括：获取目标轴承的故障信号；目标轴承设置于多级离心泵中；对故障信号进行变分模态分解，得到故障信号的分量信号；将分量信