Fault Diagnosis of Pressure Reducing Valve Based on CNN-LSTM

Aiming at the problems of diverse and closed fault modes, difficulty in extracting early weak fault features, and nonlinearity of fault data in the fault diagnosis of pressure reducing valves, this paper proposes a fault diagnosis model that integrates Convolutional Neural Network (CNN) and Long-Short Term Memory (LSTM). This model utilizes CNN to extract the spatial local features of signals and LSTM to capture the long-term dependencies in time series, effectively achieving deep extraction and precise classification of fault features. Taking the Thomas PPCD 06 pressure reducing valve as an example, the diagnosis verification is conducted for four typical faults: coil open circuit, fatigue of the reset spring, valve core wear, and valve core jamming. The experimental results show that the proposed CNN-LSTM model can effectively identify the fault states of pressure reducing valves, with an overall diagnostic accuracy rate of 98.3%.