| Manufacturing industry has come to Mass Customization production because of the intense market competition and individual customer demand. The purpose of mass customization production is quick response and diversification. Quick request for quotation has become the key process of realizing the quick response in mass customization production and how to estimate the customization product quickly has been the bottleneck of quick RFQ in mass customization. Therefore, this paper took quick cost estimation as the main research direction and method as the core content in mass customization producton, synthetically used Artificial Inteligence and Data Mining, attempted to make the new contributions to the research of quick response method in mass customization production.The manufacturing industry has already realized the importance of cost estimation methods, and the foreign and domestic research has also yielded substantial results.But the former cost estimation methods has been out of date in mass customzaton producton.Although the precision of detailed cost estimatng methods were high, their development would cost lots of financial resource, physical resource and time.Obviously this did not suit the small-batch and multi- varieties mass customization production.The development cost of quick cost estimation methods was low and process of them was quick, but their precision was unsatisfactory. So this paper proposed research construct and estimation method, and then did the case simulation to verify the cost estimation method.Specifically this paper analyzed elementary theory of product family and its modeling, then proposed product family structure management model based on GBOM, expatiated the elementary theory, principle and method of product configuration and put forward the model of creating customization product SBOM. This paper used CBR for the whole customization product, selected the most similar case and got the cost, then used adaptive ant-wavelet network to estimate the customization component, finally carried on the partial adjustment according to the formula and got the cost of customization product.This paper finally took three customization casters in Hunan Ke Hong Caster Ltd. as the simulation object, got the best parameter of adaptive ant wavelet network through compiling select language and C language, then estimated the customizationcomponent and finally adjusted the customization product. According to the data analysis this method is not only quick, but also the error of it is low, whose average error rate is 1.75%. |
PDF Full Text Request