Contaminant Source Identification in Water Distribution Networks under the conditions of Uncertainty

Optimization;Simulation;Water Distribution Systems;Identification;Contamination

Vankayala, Praveen ; Dr. Sankar Arumugam, Committee Member,Dr. S. Ranji Ranjithan, Committee Member,Dr. G. Mahinthakumar, Committee Chair,Vankayala, Praveen ; Dr. Sankar Arumugam ; Committee Member ; Dr. S. Ranji Ranjithan ; Committee Member ; Dr. G. Mahinthakumar ; Committee Chair

University:North Carolina State University

关键词: Optimization;    Simulation;    Water Distribution Systems;    Identification;    Contamination;   

Others  :  https://repository.lib.ncsu.edu/bitstream/handle/1840.16/603/etd.pdf?sequence=1&isAllowed=y

美国|英语

来源: null

【 摘 要 】

Recent attacks on the national water infrastructure have led to growing need for improved management and control of municipal water networks. Water distributions systems are susceptible to accidental and intentional chemical or biological contamination that could result in adverse health impact to the consumers.This study focuses on the contaminant source identification problem for a water distribution system under uncertainty in demand data. Due to inherent variability in water consumption levels, demands at consumer nodes in a water distribution system remain one of the major sources of uncertainty in source identification problem. In this research, the nodal demands are considered to be stochastic in nature and are varied using Gaussian and Auto Regressive models. The source identification problem is solved using the simulation-optimization method where EPANET water distribution system model acts as a simulator. Contaminant concentration observations are synthetically generated at arbitrarily selected sensor locations by specifying the location and mass loading of a potential contaminant source that is arbitrarily located at one of nodes in the system. Genetic Algorithm (GA) is used as the optimization model with the goal of finding the source location and concentration, by minimizing the difference between the simulated and observed concentrations at the sensor nodes. Two variations of GA, stochastic GA and noisy GA are applied to the same problem for comparison. Results show that noisy GA is robust and is less computationally expensive than stochastic GA in identifying the contaminant source location and concentration for both of stochastic demand models.