The Development And Research Of Intelligent Interactive Electronic Technical Manual System

Interactive Electronic Technical Manual(IETM)is a technical means to informationize the technical data of equipment,and is also one of the key technologies of military equipment support.It overcomes many disadvantages caused by the use of traditional paper technical manual,and greatly improves the work efficiency of equipment support,so it has been widely used in military and civilian fields.At present,with the change of demand in the field of equipment support and the emergence of new technologies,the development of intelligent IETM system combined with new technical means has gradually become one of the main research directions of IETM.This thesis was to complete the design and implementation of intelligent IETM system on the basis of the existing research results of IETM,the methods and key technologies for realizing intelligent system functions were researched.Firstly,according to the research status of IETM,the S1000 D standard was selected as the standard specification to be followed in IETM development,and the two core concepts of the S1000 D standard,common source database and data module,were mainly studied.Through the research of XML technology,3D model technology and other related technologies,it provided technical support for the function implementation of IETM system.Then combined with IETM user demand analysis to complete the functional design of intelligent IETM system,and improved the specific operation process of each functional module.Secondly,the intelligent extension of the retrieval function module in the system was accomplished by constructing two intelligent retrieval methods: tag retrieval and semantic retrieval,and the relevant technical methods of intelligent retrieval were researched.Tag retrieval completes the deep mining of text information through text automatic classification,built a Text CNN neural network and verified the influence of different training methods on classification results through experiments,the optimal method was selected for model training and this model was applied to the classification task of tag retrieval.Semantic retrieval was used to make up for the shortage of accurate retrieval,text vectorization was completed by word2 vec and the cosine similarity was used to evaluate semantic similarity,then the experiment proved that the method of keyword extension was more helpful to improve the accuracy of semantic retrieval.Finally,based on the data of a certain type of power station,the generation of all kinds of information objects in the IETM system was completed.In combination with the design scheme of intelligent IETM and relevant research results of intelligent retrieval,each function module was realized,so as to finally complete the construction of intelligent IETM system.