Weak faults diagnosis of high-pressure common rail systems

In order to address the challenges associated with diagnosing weak faults in high-pressure common rail systems, this study proposes a weak fault diagnosis method that is based on the Pied Kingfisher Optimizer (PKO), Variational Mode Decomposition (VMD), and Long Short-Term Memory (LSTM) networks. A fitness function is constructed by combining the cross-correlation coefficient with the cross-entropy loss function to optimize the parameters of VMD and LSTM. A fault diagnosis model for high-pressure common rail systems is established. The efficiency of the proposed model is substantiated through the implementation of an experimental platform designed for high-pressure common rail fault diagnosis. The experimental results demonstrate the efficiency of the proposed method in diagnosing plunger cracks and leaks in high-pressure common rail systems, with an accuracy rate exceeding 95%. This indicates the method's high performance in identifying and diagnosing weak faults in high-pressure common rail systems.