Prediction of cosmetic emulsion stability: Testing and mathematical modelling
Küçük Resim Yok
Tarih
2013
Yazarlar
Dergi Başlığı
Dergi ISSN
Cilt Başlığı
Yayıncı
Ege Üniversitesi
Erişim Hakkı
info:eu-repo/semantics/openAccess
Özet
Bu çalışmanın amacı, kozmetik formülasyonlarında emülsiyon stabilitesini öngörmek için bir fuzzy inference sistemi geliştirmektir. Hazırlanan on sekiz emülsiyon formülünde kıvam verici ajan, yavaş yayılan nemlendirici ve koruyucu sistem, çevresel koşullar ve depolama süresi manipüle edilen değişkenler olarak seçilmiştir. Koku ve emülsiyon stabilitesinin yanısıra, pH, viskozite ve elektriksel iletkenlik de ölçülen değerler olarak gözlemlenmiştir. Manipüle edilen ve ölçülen değişkenler arasındaki ilişki ise exhaustive search?ü takiben Adaptive NeuroFuzzy Inference System (ANFIS) kullanılarak aranmıştır.MATLAB ile geliştirilen kodların yardımıyla, manipüle değişkenlerin ölçülenler üzerine etkisi, manipüle değişkenler arasındaki etkileşimler (sinerji ya da antagonizm) ve ölçülen değişkenler arasındaki ilişki kullanıcı dostu grafik arayüzler kullanılarak incelendi.Tüm ilişkiler için öğrenme hatalarının minimum olduğu iki girişli ANFIS modeli ortaya çıkmıştır.Elde edilen sonuçlara göre, ?zaman? ve ?yavaş yayılan nemlendirici?nin beraber pH üzerine sinerjik bir etkisi vardır. Diğer ölçülen değişkenler olan viskozite, elektiriksel iletkenlik ve emülsiyon stabilitesi üzerine de ?zaman? ve ?çevresel koşullar? en etkili ikili kombinasyon olarak bulunmuştur. Ancak, koku stabilitesi için ANFIS tarafından herhangi bir manipüle değişkenlerden oluşan bir kombinasyon bulunamamıştır.viÖlçülen değerlerin stabilite parametreleriyle olan ilişkisi incelenmesine rağmen koku stabilitesi için deneysel gözlemlerle uyumlu olarak başarılı bir sonuç elde edilemedi. Ancak, emülsiyon stabilitesi üzerine ?viskozite-elektiriksel iletkenlik? çifti en etkili kombinasyon olarak bulundu.Stabilite parametreleri hariç tüm değişkenler dikkate alındığında ise, emülsiyon stabilitesi için ?viskozite-elektiriksel iletkenlik? çifti en etkili kombinasyon olarak bulunurken, koku stabilitesi için hiçbir kombinasyon bulunamadı.Anahtar kelimeler: Kozmetik, Stabilite, ANFIS, Exhaustive Search
The aim of this study is to develop a fuzzy inference system to predict the emulsion stability of cosmetic formulations. In the formulation of eighteen emulsions manipulated variables are chosen as consistency giving factor (CGF), slow spreading emollient (SSE), preservative system, environmental conditions and storage time. Together with the odor and emulsion stability pH, viscosity, and electrical conductivity are monitored as measured variables. The relationships between manipulated and measured variables are sought by exhaustive search followed by Adaptive Neuro-Fuzzy Inference System (ANFIS). With the help of codes developed in MATLAB, effects of manipulated variables on measured ones, interactions (synergy or antagonism) among manipulated variables and interrelations among measured variables are examined through user friendly graphical user interfaces. Two-input ANFIS model was arisen for all the relationships due to the minimum values of training errors. The results of indicate that ?time period? and ?SSE? together forms the synergistic combination with dominant effect on pH. For the measured variables viscosity, electrical conductivity and stability, ?time period? and ?environmental conditions? together was the most influential binary combination. But for odor stability, any combination of manipulated variables could not be found by ANFIS.Although the interrelation of output variables on stability parameters is tried to be established, no successful result is obtained for odor stability in viiiaccordance with the experimental observation. However, for emulsion stability, ?viscosity-conductivity? pair is the most effective one. All the variables except for stability parameters are considered, while ?viscosity-conductivity? is found for emulsion stability, no pair is obtained for odor stability.Keywords: Cosmetics, Stability, ANFIS, Exhaustive Search
The aim of this study is to develop a fuzzy inference system to predict the emulsion stability of cosmetic formulations. In the formulation of eighteen emulsions manipulated variables are chosen as consistency giving factor (CGF), slow spreading emollient (SSE), preservative system, environmental conditions and storage time. Together with the odor and emulsion stability pH, viscosity, and electrical conductivity are monitored as measured variables. The relationships between manipulated and measured variables are sought by exhaustive search followed by Adaptive Neuro-Fuzzy Inference System (ANFIS). With the help of codes developed in MATLAB, effects of manipulated variables on measured ones, interactions (synergy or antagonism) among manipulated variables and interrelations among measured variables are examined through user friendly graphical user interfaces. Two-input ANFIS model was arisen for all the relationships due to the minimum values of training errors. The results of indicate that ?time period? and ?SSE? together forms the synergistic combination with dominant effect on pH. For the measured variables viscosity, electrical conductivity and stability, ?time period? and ?environmental conditions? together was the most influential binary combination. But for odor stability, any combination of manipulated variables could not be found by ANFIS.Although the interrelation of output variables on stability parameters is tried to be established, no successful result is obtained for odor stability in viiiaccordance with the experimental observation. However, for emulsion stability, ?viscosity-conductivity? pair is the most effective one. All the variables except for stability parameters are considered, while ?viscosity-conductivity? is found for emulsion stability, no pair is obtained for odor stability.Keywords: Cosmetics, Stability, ANFIS, Exhaustive Search
Açıklama
Anahtar Kelimeler
Kimya Mühendisliği, Chemical Engineering
