An Algorithm Based On Hybrid Self-adaptive Ant Colony Algorithm For Product Configuration Under Mass Customization

Along with continuous development of manufacturing technology，MassCustomization mode has become the inevitable style with the development ofmanufacturing industry. In order to adapt to the continuous improvement ofthe production scale and product complexity，Product Configuration as thekey technology of Mass Customization production，the research of algorithmfor Product Configuration has become the focus of attention. As a result，AnAnalytic Hierarchy Process（AHP）is adopted to propose a multi-objectiveproduct configuration model，and an algorithm based on Hybrid Self-adaptiveAnt Colony Algorithm（ACA） for Product Configuration is provided toresolve product configuration effectively under Mass Customization in thispaper.Firstly， the core concept of Mass Customization and ProductConfiguration are summarized， the typical algorithms in ProductConfiguration are introduced systematically. We also categorized them in the paper. In ant societies，the activities of the individuals，as well as of thesociety as a whole，are not regulated by any explicit form of centralizedcontrol. The presence of these has made ant societies an attractive andinspiring model for building Ant Colony Algorithm. Ant Colony Algorithmconverges on the optimization path through information pheromoneaccumulation. It is one of relatively good practical intelligent algorithms. Theresearch on experimental analysis of the effect of pertinent parameters in AntColony Algorithm is focused in the paper.Then，for the multi-objective optimization problem based on productmodular design，an fuzzy Analytic Hierarchy Process is adopted to determinethe weights of sub-objective functions.At last，based on the study of Ant Colony Algorithm，the thesis presentsan algorithm based on Hybrid Self-adaptive Ant Colony Algorithm forproduct configuration. Theoretical analysis show that the algorithm canmaintain advantages that Ant Colony Algorithm can give the precision of thesolution. Simultaneously the algorithm further achieved a better balance in theglobal search ability and convergence speed.