化粧品処方へのバーチャルパラメータ設計の応用

抄録

A study was made of the usefulness of introducing virtual parameter design into the technology design process, at the stage following system selection. It was thought that results by virtual parameter design would make it possible to decide whether to proceed to actual parameter design. When virtual parameter design was applied to a cosmetic formulation that had previously been designed by conventional methods, different evaluators assessed the formulation differently, and the results suggested that inappropriate noise factors had been used. After reconsideration of the setting of factors and the method of putting questions, a second virtual parameter design was carried out, and greater reproducibility was observed than in the first virtual parameter design. Next a confirmation experiment was performed, using actual formulations. A certain degree of reproducibility was obtained in the S/N ratio of the error variance, and it was found that an overall picture of the cosmetic had been obtained, but it was also fbund that signal and noise factors had not been distinguished.