Aviation Plunger Pump Fault Prediction and Diagnosis Method Based on the Fusion of Knowledge Graph
Aiming at the problems of high fault frequency,many fault types,difficult fault traceability and low prediction accuracy in the operation of aviation plunger pump,a fault prediction and diagnosis method of aviation plunger pump based on the fusion of knowledge graph and artificial bee colony algorithm is proposed.Firstly,the knowledge graph architecture,entity types and inter-entity relationships are defined from top to bottom.The knowledge network of the graph is constructed from bottom to top,and entity naming recognition,extraction,fusion,integration and reasoning are performed for the data layer.Secondly,the artificial bee colony fault prediction algorithm is established,which includes input layer,assignment layer,propagation layer,self-attention layer and output layer.The fault feature extraction,variable neighborhood bidirectional gating fault prediction and attention mechanism are used to form the aviation plunger pump fault prediction model through feature vector training.Finally,through the actual maintenance cases,the knowledge graph of aviation plunger pump fault diagnosis is constructed.The experiment proves the effectiveness and feasibility of the above method,and verifies the efficient fault diagnosis ability of the proposed algorithm.
万方数据
维普
