Premium
A Mamdani Adaptive Neural Fuzzy Inference System for Improvement of Groundwater Vulnerability
Author(s) -
Agoubi Belgacem,
Dabbaghi Radhia,
Kharroubi Adel
Publication year - 2018
Publication title -
groundwater
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.84
H-Index - 94
eISSN - 1745-6584
pISSN - 0017-467X
DOI - 10.1111/gwat.12634
Subject(s) - adaptive neuro fuzzy inference system , hydrogeology , vulnerability (computing) , fuzzy logic , computer science , aquifer , artificial neural network , groundwater , inference system , environmental science , fuzzy inference system , water resource management , hydrology (agriculture) , machine learning , artificial intelligence , fuzzy control system , engineering , geotechnical engineering , computer security
Assessing groundwater vulnerability is an important procedure for sustainable water management. Various methods have been developed for effective assessment of groundwater vulnerability and protection. However, each method has its own conditions of use and, in practice; it is difficult to return the same results for the same site. The research conceptualized and developed an improved DRASTIC method using Mamdani Adaptive Neural Fuzzy Inference System (M‐ANFIS‐DRASTIC). DRASTIC and M‐ANFIS‐DRASTIC were applied in the Jorf aquifer, southeastern Tunisia, and results were compared. Results confirm that M‐ANFIS‐DRASTIC combined with geostatistical tools is more powerful, generated more precise vulnerability classes with very low estimation variance. Fuzzy logic has a power to produce more realistic aquifer vulnerability assessments and introduces new ways of modeling in hydrogeology using natural human language expressed by logic rules.