Font Size: a A A

Subject Recognition And Trend Prediction Of New Energy Vehicle Technology Based On Patent Data

Posted on:2022-08-25Degree:MasterType:Thesis
Country:ChinaCandidate:W Q LinFull Text:PDF
GTID:2492306539970259Subject:Management Science and Engineering
Abstract/Summary:PDF Full Text Request
With the increasingly serious problems of environmental pollution and energy crisis,the development of new energy vehicle industry has become one of the effective means to alleviate the current problems.Therefore,countries have promulgated a series of policies and measures to support the development of new energy vehicle industry,trying to plan to accelerate the rapid promotion of new energy vehicles through technological innovation,so as to reduce the dependence of the automobile industry on oil,To avoid the adverse effects of automobile exhaust on the environment and realize the sustainable development of environment and economy.Facing the increasingly fierce market competition environment,technological innovation has become an important tool for automobile manufacturing enterprises to participate in market competition.However,technological innovation is not a blind process,the first step in the implementation of innovation is the need for technology discovery,clear direction and path of technological innovation.This is a crucial step in technological innovation,which is related to the success of technological innovation.It is the main force of new energy vehicle technology innovation,facing the increasingly fierce market competition environment,therefore,only grasp the direction of technology development,identify technological innovation opportunities,and formulate the optimal technology development strategy in time,can it comply with the trend of rapid development of new energy vehicle technology.Patent literature,as the material of recording technological innovation,hides the important information of technological innovation.Based on patent literature,text mining research on new energy vehicle technology patent literature can quickly understand the distribution of technology topics,analyze the key points of technology research and development,and predict the trend of technology development,which is of great significance for the government or enterprise technology innovation management.This paper takes new energy vehicles as the research object,takes patent documents collected from CNKI patent database as the data basis,combines text mining,bibliometrics and patent analysis related methods,and takes "theoretical methods and literature review →patent document subject identification → hot technology subject trend analysis →technology subject importance trend analysis" as the main line of this paper,puts forward a set of new energy vehicles patent analysis methods The main innovation achievements are summarized as follows:(1)Subject recognition of new energy vehicle technology based on patent documentsA patent text mining system framework based on topic model this paper introduces topic model into the field of patent text mining and science and technology management,develops and realizes a complete process from data input,text cleaning,topic recognition and topic analysis,so as to realize the technology topic recognition of massive new energy vehicle patent text and find the important technology topics.(2)Building Arima prediction model based on document subject distributionBy combining the document topic distribution of topic model training with time series,the development trend of technology topic research heat can be preliminarily fitted.The top5 technology topics are selected as the object of ARIMA model construction,and the heat trend is predicted.(3)Constructing HMM prediction model based on co-occurrence of keywordsThe initial parameters of HMM model are constructed based on the topic words of topic model,and the double random hidden Markov model is constructed to predict and analyze the importance and development trend of technology topic.
Keywords/Search Tags:topic model, new energy vehicle, Trend forecast, ARIMA model, HMM model
PDF Full Text Request
Related items