Skoči na glavni sadržaj

Izvorni znanstveni članak

https://doi.org/10.30765/er.40.1.07

Estimation Of Los Angeles And Micro Deval Abrasion Values For Gneiss And Granite Using Anfis Approach

Irina Pachoukova ; Department of Civil Engineering, Ecole National Superieure d'Ingenieurs, University of Lome, Lome, Togo
Adekunlé A. Salami orcid id orcid.org/0000-0002-0917-1231 ; Department of Electrical Engineering, Ecole National Superieure d'Ingenieurs, University of Lome, Lome, Togo
Ayité S. A. Ajavon ; Department of Electrical Engineering, Ecole National Superieure d'Ingenieurs, University of Lome, Lome, Togo
Ampah Kodjo Christophe Johnson ; Geological Department, Faculty of Scienece, University of Lome, Lome, Togo


Puni tekst: engleski PDF 578 Kb

str. 48-58

preuzimanja: 500

citiraj


Sažetak

Gneisses and granites mechanical properties knowledge such as the Los Angeles and Micro-Deval coefficients is very important for the engineer when designing and building civil engineering works. Laboratories test often performed to obtain Los Angeles and Micro-Deval values are expensive and time consuming. For this, the depths (parallel and perpendicular to the foliation planes) obtained during a rock drilling test will be used to estimate its coefficients Los Angeles and Micro-Deval. In this study, an ANFIS approach is used to estimate gneisses and granites Los Angeles and Micro-Deval that have been compared to the Multiple Linear Regression method. For this purpose, we have used a database of a sample size of 80 to determine the parameters of the ANFIS and Multiple Linear Regression   models, and a second sample of size 15 used for the validation tests. The results obtained show us that we can estimate the mechanical properties (Los Angeles and Micro-Deval) of gneisses and granites with ANFIS approach.

Ključne riječi

prediction; ANFIS; MLR; Los Angeles abrasion test (LA), Micro-Deval abrasion test (MD); rock drilling test

Hrčak ID:

232826

URI

https://hrcak.srce.hr/232826

Datum izdavanja:

27.1.2020.

Posjeta: 1.431 *