Approximate life cycle assessment of classified products usingartificial neural network and statistical analysis in conceptual productdesign

摘要

In the early phases of the product life cycle, Life CycleAssessment (LCA) has been used to support decision-making for conceptualproduct design; the best alternative can be selected based on itsestimated LCA and its benefits. Both the lack of detailed informationand time for a full LCA for a various range of design conceptsdemonstrate the need for a new approach to environmental analysis. Thepaper suggests a novel approximate LCA methodology for the conceptualdesign stage by grouping products according to their environmentalcharacteristics and by mapping product attributes to impact driverindex. The relationship is statistically verified by exploring thecorrelation between total impact indicator and energy impact category. Aneural network approach is developed to predict an approximate LCA ofgrouping products in conceptual design. Trained learning algorithms forthe known characteristics of existing products will quickly give theresult of LCA for new design products. Training is generalized by usingproduct attributes for an ID in a group as well as other productattributes for other IDs in other groups. The neural network model withback propagation algorithm is used and the results are compared withthose of multiple regression analysis. The proposed approach does notreplace full LCA but it provides some useful guidelines for the designof environmentally conscious products in the conceptual design phase