Considering the complicated failure causes of aircraft manufacturing process

the description has the features of long and complex text. Based on the deep learning theory and knowledge engineering

a fault intelligent diagnosis was proposed for the aircraft manufacturing process. Firstly

the attribute reduction of the quality failure form was carried out. The word segmentation approach was utilized to reduce the long text attribute on basis of the thesaurus and full-model participle. Secondly

the long text attribute was vectorized by the deep learning algorithm. The k-nearest neighbor (KNN) and the cosine similarity algorithm were investigated to retrieve the similar forms from the historical fault forms for aircraft fault diagnosis. Finally

the fault intelligent diagnosis method was validated by the failure case of a certain aircraft manufacturing process. The results show that the proposed method can help the aircraft maintenance personnelmake dispositions

reduce the work loads

and improve maintenance efficiency.