Research On Customer Requirements Discovery And Product Optimized Configuration Based On Automobile Big Data

Automobile product development engineering is a practical process in which the developers use engineering tools,techniques and methods to plan and optimize the resources of human,financial,material,energy and information,and to transform market requirement into products.It is the main force for the realization of automobile product innovation.Customer requirement acquisition is the basis of automobile product development process such as automobile product concept design,competitive product analysis and establishment of product specification.Acquiring and meeting customer needs is the primary concern for automobile product development,and transforming customer requirement into automobile product configuration is an important process in automobile product development.Focusing on the acquisition of customer requirement,people summarized the methods of requirement discovery such as questionnaire survey,interview,focus group discussion,and empathy design.With the changes of market environment and the improvement of people's consumption level,the personalization,fragmentation and variability of customer requirement information have brought new problems to requirement acquisition.The customer requirements acquired by conventional methods are not complete,objective,and comprehensive,and are not sufficient to support a customer-centric automobile product development strategy.In the Internet and big data environment,the development of social media technology and smart internet technology provides a new way for the expression and collection of customer requirement information.Based on the systematic analysis of related theories and methods,this paper uses automobile review data and operation data to deeply study the key issues of automobile product requirement acquisition,requirement information expression,decomposition,mapping and product reconfiguration.A customer requirement identification process based on automobile review data is established,which has solved the problem of acquiring automobile product requirement from social media big data.A customer requirement identification process based on electric vehicle operation data is established,which has solved the problem of acquiring automobile product requirement from electric vehicle operation data.Based on the customer requirement acquired from the automobile big data,and considering the dynamics of the customer requirement,the processing of the requirement information and the method of product reconfiguration after customer requirement change are proposed.The specific research work and innovative research results of this paper are as follows:(1)The customer requirement identification method based on automobile online reviews is studied.The product features in the automobile review data are extracted.The SVM model is used to identify the spam comments.The “feature-view” pairs in the product reviews are identified and their emotional polarity is determined.The nonparametric exponential smoothing model Holt-Winters is used to predict the product requirement trend,and Mann-Kendall method is used to test the trend change of the attention and emotional orientation of attributes.(2)The customer requirement identification method based on electric vehicle operation data is studied.Factor analysis is used to reduce the dimension of vehicle operation data.The K-means method is used to cluster the data to obtain the driving behavior characteristics of different categories of users.Considering the sequential characteristics of user behavior,time series clustering is used to further identify the automobile usage pattern.Considering the different regional preferences of users,descriptive statistical analysis is used to identify the regional preference of users.(3)The automobile product reconfiguration method for customer requirement change is studied.The customer requirement acquisited from the automobile review data and operation data is expressed with standardized words.The concept learning system,the semantic network and the QFD method are respectively used to identify,decompose and map the requirement change information.The commonality analysis method of automobile product modules and components is proposed,and the UB-BU hybrid optimization method is used to reconfigure the general parts.Customer requirement acquisition is an important way for automobile product stakeholders to gain insight into the consumer market.Multi-channel,multi-angle and allaround customer requirement acquisition methods and reasonable product configuration process are the basis for automobile enterprises to formulate product development strategies,which is related to the success or failure of automobile product development.With the development of Internet,big data,Internet of things and other emerging information technologies,the customer requirement information source has developed from a simple text to a variety of new media forms such as sound,image and video,and the configuration process and methods have become richer.How to make better use of multi-source information and intelligent technology,to achieve the comprehensiveness,accuracy and timeliness of customer requirement information acquisition,as well as automation,digitization and intellectualization of product configuration,so as to provide decision-making support for product development and design,is an important research direction in the future.