Graphical Estimation of Permeability Using RST&NFIS

TL;DR

This paper combines Rough Set Theory, Self Organizing Maps, and Neuro-Fuzzy Inference Systems to analyze permeability data from a dam, providing a comparative approach for permeability estimation.

Contribution

It introduces a novel combination of RST, SOM, and NFIS for permeability analysis, highlighting the effectiveness of integrated fuzzy and rough set methods.

Findings

RST identified dominant rules for permeability levels.

SOM-NFIS provided comparable or improved permeability estimations.

The combined approach offers a new methodology for dam permeability analysis.

Abstract

This paper pursues some applications of Rough Set Theory (RST) and neural-fuzzy model to analysis of "lugeon data". In the manner, using Self Organizing Map (SOM) as a pre-processing the data are scaled and then the dominant rules by RST, are elicited. Based on these rules variations of permeability in the different levels of Shivashan dam, Iran has been highlighted. Then, via using a combining of SOM and an adaptive Neuro-Fuzzy Inference System (NFIS) another analysis on the data was carried out. Finally, a brief comparison between the obtained results of RST and SOM-NFIS (briefly SONFIS) has been rendered.