Statistical analysis of the performance of the soft computing based prediction model for shrinkage of concrete including mineral admixtures

المؤلفون المشاركون

Arbili, Muhammad M.
Ghaffuri, Farman K.
Mermerdas, Qasim

المصدر

ZANCO Journal of Pure and Applied Sciences

الناشر

جامعة صلاح الدين قسم النشر العلمي

تاريخ النشر

2016-06-30

دولة النشر

العراق

عدد الصفحات

7

التخصصات الرئيسية

الهندسة المدنية

الملخص الإنجليزي

The study covers a numerical study to derive a soft computing based mathematical model for prediction of the shrinkage of concretes having different mix compositions.

The model will be derived by means of gene expression programming (GEP).

In order to generate the model, a comprehensive data set compiled from technical literature will be used.

The prediction parameter will be assigned as water-to-binder ratio (w/b), silica fume content (SF) in kg/m3, fly ash content (FA) in kg/m3, cement (C) content in kg/m3, aggregate/binder ratio, compressive strength (fc) in MPa, type of shrinkage for shrinkage (drying or autogenous shrinkage), and drying time in days.

The GEP model obtained in this study will be compared to an available formula which has been presented by first two authors of this paper.

The available model was generated by artificial neural networks (NN).

The comparison of the models will be carried out through some statistical parameters such as correlation coefficient (R), mean absolute percent error (MAPE), means square error (MSE), and root mean square error (RMSE).

Moreover, in order to assess performances of the proposed models an advanced statistical analysis method called “Wilcoxon rank sum test” was applied.

The test indicates whether the obtained prediction and experimental datasets are statistically equivalent or not at a specified level of significance.

نوع البيانات

أوراق مؤتمرات

رقم السجل

BIM-786858

• APA
• MLA
• AMA