By using neural network, grey correlation degree and fuzzy rules intelligent recognition method, respectively to known aero engine oil samples of wear particle recognition, inspection training sample number, algorithm structure and target recognition of the complexity of factors restricting, the recognition accuracy and stability of the defects, and by D-S evidence theory to the recognition results are fused, comprehensive utilization of the advantages of each model, distinguishability and accuracy of wear particle recognition, and the level of aviation oil monitoring technology based on wear particle recognition have been greatly improved.