Interactive Educational Module to Enhance Understanding of Genetic Algorithms in Surface Water Resources Systems Analysis

An educational module has been developed to introduce graduate-level students to genetic algorithms and their applications in surface water systems analysis. The module includes a user-friendly graphical user interface that is used to select genetic algorithm heuristics and parameters and explore their effect on the analysis of an irrigation canal network management problem. The simulation model for the canal network is an unsteady flow model that predicts the hydraulic response in the network to changes in boundary conditions and internal sluice gate operations, as well as the irrigation demands in the areas served by the network. This model is incorporated into a genetic algorithm-based search framework to find efficient strategies for scheduling gate operations and boundary conditions that minimize the amount of water consumed while satisfying constraints throughout the network (related to flooding, water shortages, stability of gates, etc.) over a period of time. This module has been developed primarily for use in graduate level water resources systems analysis courses, and is intended to help students visualize and understand the effects of different GA operators, parameters and constraint-handling techniques on the search process and the solution quality.