| Product family(PF)strategy is the main way to achieve mass customization,aiming to meet the personalized needs of customers while simultaneously possessing the advantages of high efficiency and low cost of mass production mode.Studying intelligent design methods for PF can not only reduce the complexity of issues in the PF development process,but also facilitate the industrial upgrading and intelligent manufacturing of enterprises.At present,the research on intelligent PF design methods in domestic and foreign literature is relatively limited and lacks systematicness.Therefore,aim at the PF design process,this thesis has researched the theory and method of intelligent design for PF.The main research contents are as follows:(1)The research status of intelligent design methods for PF in domestic and foreign literature is reviewed,so the shortcomings of previous research are analyzed.Moreover,the basic theories of PF design and intelligent design are summarized,then the sources,types,mining methods,and application scenarios of PF data knowledge are concluded.Based on this,the research framework of this thesis is put forward.(2)Systematic and intelligent design methods for PF architecture and configuration have been proposed.Firstly,the principal component analysis(PCA)algorithm is used to reduce the dimension and simplify the customer need data,aiming to extract the customer’s differentiated needs.Then,association rule algorithm is used to mine configuration rules between customer differentiated needs and module instances,so a Classifier Based on Association Rule(CBA)is constructed based on the mined rules.Secondly,sequence comparison technology is applied for module identification,so as to build the product platform and PF architecture.According to the personalized needs of customers,designers use CBA classifiers to output the best product configuration scheme,and configure products based on the PF architecture.Finally,the effectiveness of the proposed method is verified with the desktop computer host PF,and the results show that the proposed method improves the accuracy and intelligence of product configuration compared to traditional methods.(3)An intelligent optimization method for supplier selection under a carbon neutral policy is studied.Firstly,the ontology representation method is used to establish the data model of PF manufacturer and suppliers.By calculating the semantic similarity between the supplier ontologies and the manufacturer ontology,candidate suppliers that meet the manufacturer’s needs are initially selected.Secondly,based on data of module instances provided by candidate suppliers,the carbon footprint of the PF is quantified.Then,with the goal of minimizing the manufacturer’s total cost and the constraints of carbon neutral policy,an optimization model for the PF configuration cost-carbon neutral cost(CC-CNC)is constructed,and then genetic algorithm is empoyed to solve the model to mine the optimal supplier selection scheme.Finally,the effectiveness of the proposed model was verified by the case study of water cooled fan PF.Moreover,based on the case study,the impact of carbon neutral policy on manufacturers’ decisionmaking was analyzed,and the analysis results are used to guide manufacturers to take reasonable carbon neutral measures under carbon neutral policy.(4)An intelligent identification method for the priority of PF redesign is put forward.Firstly,web crawlers,text recognition,and emotional analysis techniques are utilized to model customer online reviews into structured datasets,thereby identifying the PF performances that customers focus on in online reviews,and quantifying customer satisfaction.Then,calculate the importance factor,commonality factor,satisfaction factor,and Kano factor for each performance specification of PF,and then the redesign priority for each performance is quantified based on these four factors,thereby guiding enterprises to redesign the PF performances with limited costs and resources.Finally,the effectiveness of the proposed method is verified by the research example of a sweeping robot PF.Compared with traditional methods,the proposed method can more efficiently and comprehensively mine the customer opinions,as well as the focus of product family redesign.(5)Develop software system for PF intelligent design.Based on the theory and method of PF intelligent design proposed in this thesis,a PF intelligent design software is developed.Firstly,the main functions of the software system are analyzed,and the overall architecture of the software system is planned based on functional requirements.Then,the PF intelligent design software is developed based on the Microsoft Visual Studio development platform,Visual Basic programming language,and My SQL database.Finally,combined with research cases,the main functional modules and operating procedures of the software system are introduced in detail.At the end of the thesis,a summary of this study is presented and further research work is discussed.There are 59 figures,26 tables,and 198 references in the thesis. |
