Using Niched Co-Evolution Strategies to Address Non-Uniqueness in Characterizing Sources of Contamination in a Water Distribution System

Integrating evolution strategies and genetic algorithms with agent-based modeling for flushing a contaminated water distribution system

“…On-going research is investigating development and application of the coupled ABM-evolutionary computation framework to manage threats for a larger, more realistic municipalify (Shafiee & Zechman 2ona, b, 2012;Zechman et al 2on). The framework described here demonstrates a new approach that can be taken to assist utilify operators in responding to hazardous events.…”

“…Two hydrants are considered hydraulically connected where there is no other hydrant in the shortest flow path between the pair of hydrants. This mutation operator is based on the mutation operator that was developed for identifying the source of contamination in a water distribution system (Zechman & Ranjithan 2009;Drake & Zechman 2on), and it has been applied to resize pipes in a water distribution system to meet fire flow requirements (Kanta et al 2012) and route siren vehicles during a water contamination event (Shafiee & Zechman 2onb). …”

Contaminant source identification of water distribution networks using cultural algorithm

“…Guan et al proposed that the simulation optimization method is used to solve the CSI of nonlinear problem by constantly reading sensor data optimization and correction of pollutant sources and identified that the source of pollutants and the historical time eventually released pollutants. Many other optimization methods are also used by this field . Liu et al present that the method based on evolutionary algorithm, using dynamic optimization techniques to adapt to search source characteristic (start time, location, and history) released by constantly adding new sensors, is available and the optimum solution of slow convergence is the only one .…”

“…Many other optimization methods are also used by this field. [23][24][25][26][27] Liu et al present that the method based on evolutionary algorithm, using dynamic optimization techniques to adapt to search source characteristic (start time, location, and history) released by constantly adding new sensors, is available and the optimum solution of slow convergence is the only one. 28 In the study, 29 an optimal source identification model incorporating adaptive simulated annealing optimization algorithm linked with the numerical flow and transport simulation models is designed to identify contaminant source characteristics.…”