Document Type


Publication Date



In the event that a contaminant enters a water distribution system, opening hydrants to flush contaminated water can protect consumers from becoming exposed. Strategies for operating hydrants can be developed to specify the selection of hydrants and the timing of operations to maximize the amount of contaminant that is removed from the system. As an event unfolds, however, sensor data may be the only information that is available to indicate the location and timing of the contaminant source, and ultimately, hydrant strategies must be selected in a highly uncertain environment. The decision-making framework for making real-time decisions to select hydrant strategies relies on computational and sensor technologies, including the accuracy and precision of sensor data; the timeliness of data availability (e.g., streaming data or data that is collected manually); and computational capabilities to execute search simulation-optimization frameworks in real-time. This research will explore and compare two decision-making frameworks. The first framework integrates real-time algorithms to identify potential source locations and develop hydrant strategies, using precise water quality data and high performance computation. The source identification problem is solved using a multi-population evolution strategies approach, and a genetic algorithm approach is applied to identify hydrant strategies for specified source locations. The second decision-making framework provides a library of response options that can be selected based on sensor data as an event unfolds. The library of hydrant strategies is developed a priori using a simulation-optimization framework. Potential sources are classified based on the order of sensors that are activated, and hydrant strategies are identified to maximize average performance for events within each class through the application of a genetic algorithm framework. The two decision-making frameworks are applied and compared for a set of events that are simulated for a virtual city, Mesopolis.


Session S6-03, Special Session: Evolutionary Computing in Water Resources Planning and Management III


To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.