Knowledge graph-based fan omen fault timing early warning method and system

Abstract

The application discloses a kind of fan omen fault timing early warning method and system based on knowledge graph, it is related to fan technical field.Extract the timing state feature matrix containing the operating condition characteristics of each key component of fan;Extract the real-time dynamic anomaly subgraph composed of potential abnormal nodes and their topological neighbors;Build a hybrid query prompt vector for graph retrieval;Recall the set of historical fault evolution paths that highly match the current abnormal pattern;Fusion generates a retrieval-enhanced syndrome evolution graph that carries historical fault context;Calculate the state transition probability and cross-node causal dependence weight matrix of each symptom node in the graph between consecutive time slices;Output fan symptom failure warning instruction.The present application enables the network to accurately capture the cross-component cascading deterioration trend during graph fusion, significantly enhancing the prediction accuracy of fault propagation between complex components.

Images