Artificial intelligence methods in water systems research – a literature review

Abstract

We over view se lected ar ti fi cial in tel li gence meth ods used in re search on wa ter sys tems, spe cif i cally ar ti fi cial neu ral net works (ANN), adap tive neuro-fuzzy in fer ence sys tems (ANFIS), ge netic pro gram ming (GP) and sup port vec tor ma chine (SVM) meth ods. Each method is char ac ter ized and the most ef fec tive ways of us ing these meth ods are dis cussed. These meth ods prove widely use ful in fore cast ing changes in se lected sur face and ground wa ter qual ity pa ram e ters, fore cast ing sew age network fail ures, as sess ing wa ter treat ment op tions, cli mate mon i tor ing, drought de tec tion and en vi ron men tal is sues for farmers and pro duc ers. Pub lished stud ies show that ar ti fi cial in tel li gence meth ods should be used in the anal y sis of wa ter sys tems, es pe cially since ar ti fi cial in tel li gence now ap pears in search re sults for over 60,000 en vi ron men tal ar ti cles. Key words: hydrogeology, ma chine learn ing, ur ban wa ter sys tems, ar ti fi cial in tel li gence, sen sors, anal y sis.