Integrated Decision Support System for Prognostic and Diagnostic Analyses of Water Distribution System Failures

Abstract

This paper presents an innovative decision support system (DSS) for prognostic and diagnostic analyses of water distribution system (WDS) failures. The framework of the DSS is based on four novel models developed and published by the authors of this paper. The four models include reliability assessment model, leakage potential model, leakage detection model, and water quality failure potential model. Information obtained from these models together with external information such as customer complaints, lab test results (if any), and historical information are integrated using Dempster-Shafer (D-S) theory to evaluate prognostic and diagnostic capabilities of the DSS. The prognostic capabilities of the DSS provide hydraulic and water quality states of a WDS whereas the diagnostic capabilities of the DSS help to identify the failure location with minimal time after the occurrence and will help to reduce false positive and false negative predictions. The framework has ‘unique’ capacity to bring the modeling information (hydraulic and Quality), consumer complaints, historical failure data, and laboratory test information under a single platform to perform a prognostic and diagnostic investigation of WDS failures (hydraulic and Quality). The proof of concept of the DSS has been demonstrated using data used in published four articles. The outcomes of this research widely addressed the uncertainties associated with WDS which improves the efficiency and effectiveness of diagnosis and prognosis analyses of WDS. It is expected that the developed integrated framework will help municipalities to make informed decisions to increase the safety, reliability and the security of public health.