Multiobjective genetic algorithm conjunctive use optimization for production, cost, and energy with dynamic return flow

作者：Peralta Richard C^*; Forghani Ali; Fayad Hala

来源：Journal of Hydrology, 2014, 511: 776-785.

DOI：10.1016/j.jhydrol.2014.01.044

• Conjunctive use
• Genetic algorithm
• Multiobjective optimization
• Nonlinear aquifer

Many real water resources optimization problems involve conflicting objectives for which the main goal is to find a set of optimal solutions on, or near to the Pareto front. E-constraint and weighting multiobjective optimization techniques have shortcomings, especially as the number of objectives increases. Multiobjective Genetic Algorithms (MGA) have been previously proposed to overcome these difficulties. Here, an MGA derives a set of optimal solutions for multiobjective multiuser conjunctive use of reservoir, stream, and (un)confined groundwater resources. The proposed methodology is applied to a hydraulically and economically nonlinear system in which all significant flows, including stream-aquifer-reservoirdiversion-return flow interactions, are simulated and optimized simultaneously for multiple periods. Neural networks represent constrained state variables. The addressed objectives that can be optimized simultaneously in the coupled simulation-optimization model are: (1) maximizing water provided from sources, (2) maximizing hydropower production, and (3) minimizing operation costs of transporting water from sources to destinations. Results show the efficiency of multiobjective genetic algorithms for generating Pareto optimal sets for complex nonlinear multiobjective optimization problems.

• 出版日期2014-4-16