Sensor Placement in Municipal Water Networks

Abstract

We present a model for optimizing the placement of sensors in municipal water networks to detect maliciously injected contaminants. An optimal sensor configuration minimizes the expected fraction of the population at risk. We formulate this problem as a mixed-integer program, which can be solved with generally available solvers. We find optimal sensor placements for three test networks with synthetic risk and population data. Our experiments illustrate that this formulation can be solved relatively quickly, and that the predicted sensor configuration is relatively insensitive to uncertainties in the data used for prediction.

Stochastic sampling design using a multi-objective genetic algorithm and adaptive neural networks

“…In water quality management, optimal sensor placement in WDS has also attracted special attention with the aim of identifying contamination sources (Ostfeld and Salomons 2004;Berry et al 2005;Propato, 2006;Berry et al 2006;Shastri1 and Diwekar 2006). They all typically minimize the risk from contamination using sensors for timely detection.…”

Analyzing the Vulnerability of Critical Infrastructure to Attack and Planning Defenses

“…We will not provide details of this min-max defender-attacker model, but we note that related models have been studied for detecting malevolent contamination of a municipal water system (e.g., Berry et al [5]). Figure 4 shows the value of the optimal solution for varying numbers of detectors.…”

Incremental Sensor Placement Optimization on Water Network

“…[4,5] introduced a scenario in which the objective is to ensure a pre-specified maximum volume of contaminated water consumed prior to detection and also reduced this problem to set cover problem. [6,7] introduced an MIP(mixed integer programming) solution for the objective to minimize the expected fraction of population exposed to a contamination. The objective of [8,9] is to ensure that the expected impact of a contamination event is within a pre-specified level, and [8] introduced a formulation based on set cover and solved the problem using genetic algorithm while [9] use a MIP based solution.…”

“…We start this section with the introduction of preliminary concepts in Sec 3.1. Please refer to [6] for more background knowledge. Then, we propose a scenario-cover based model in Sec 3.3, upon which ISP is defined in Sec 3.2.…”

“…More specifically, d c,v represents the total damaging influence caused by contamination scenario c during the time period from the beginning of c to the time point at which c was detected by a sensor deployed at vertex v. d c,v can be defined from various aspects, including volume of contaminated water consumed [4], population affected [6], and the time until detection [2]. In this paper, we use the time until detection as the quantitative criteria to evaluate the adverse impact of each contamination scenario.…”