Research on Automatic Engine Fault Diagnosis Based on Multi Information Fusion Feature
The complexity of the engine process and the complexity of its fault diagnosis informa-tion make it difficult to identify and diagnose fault problems with a single sensor.The study therefore proposes a multi-information fusion of the data,feature and diagnostic layers,and proposes a diagnos-tic model based on maximum information features and principal component analysis.The experimental results show that the model has a deviation value of less than 5％for data extraction of different fault types,and its error performance(＜0.06％)and testing accuracy are significantly better than other comparative algorithms.The multi-information fusion feature diagnosis algorithm can effectively diag-nose faults in engine equipment and provide new maintenance ideas.
multi information fusionenginefault diagnosisautomationfeature recognitionD-S evidence theory
万方数据
维普
