Application of Novel MCDM for Location Selection of Surface Water Treatment Plant

Abstract

Surface water treatment plants (SWTPs) are responsible for supplying treated water to urban or rural consumers to satisfy the demand of the proximal population for drinking water. One important reason why an SWTP may fail is the poor selection of its location. This article proposes an automated framework for decision making which was developed empirically to identify a location objectively and cognitively, with consideration of all relevant indicators, selected in relation to their role in ensuring the optimality of SWTP performance. Thus, a new multicriteria decision making method was developed to identify the most common and significant indicators and to develop an index that would capture the feasibility of a given location for the installation of an SWTP. In addition to this method, another novel predictive model was developed, based on a polynomial neural network architecture and bagged modeling, to automate the process of the feasibility assessment of the location. These two models were developed and utilized for the first time for the automatic assessment of location feasibility for an SWTP installation. The prototype application of the two new models concerned a few test locations of a peri-urban metro city, located in northeast India, where the results obtained from the model seconded the real scenario for the test locations. The results encourage further application of this process.