Abstract
Competitive optimization techniques have been developed to address the complexity of integrated water resources management (IWRM) modelling; however, model adaptation due to changing environments is still a challenge. In this paper we employ multi-variable techniques to increase confidence in model-driven decision-making scenarios. Here, water reservoir management was assessed using two evolutionary algorithm (EA) techniques, the epsilon-dominance-driven self-adaptive evolutionary algorithm (-DSEA) and the Borg multi-objective evolutionary algorithm (MOEA). Many objective scenarios were evaluated to manage flood risk, hydropower generation, water supply, and release sequences over three decades. Computationally, the -DSEA’s results are generally reliable, robust, effective and efficient when compared directly with the Borg MOEA but both provide decision support model outputs of value.
Highlights
Water resource management problems are complex due to their non-linear, dynamic, multimodal properties that need robust methods to solve, such as optimization algorithms [1] based on evolutionary algorithms (EAs) inspired from evolution and the natural selection of species [2,3]
Many of these have been proposed by researchers with different techniques, such as: the non-dominated sorting genetic algorithm (NSGA II) [4], multi-objective evolutionary algorithm based on decomposition (MOEA/D) [5], indicator-based evolutionary algorithm (IBEA) [6]
A review of EAs and other metaheuristic algorithms and their applications can be found in [11,12]. Examples using these techniques for solving water resources management problems include Hurford et al, [13], and others [14,15,16,17] using ε-NSGA-II, MOEA/D, Borg MOEA and NSGA-II, respectively to optimize reservoir management strategy based on multidisciplinary objectives like flood control, hydropower generation, and water supply
Summary
Water resource management problems (i.e., surface and groundwater) are complex due to their non-linear, dynamic, multimodal properties that need robust methods to solve, such as optimization algorithms [1] based on evolutionary algorithms (EAs) inspired from evolution and the natural selection of species [2,3]. Examples using these techniques for solving water resources management problems include Hurford et al, [13], and others [14,15,16,17] using ε-NSGA-II, MOEA/D, Borg MOEA and NSGA-II, respectively to optimize reservoir management strategy based on multidisciplinary objectives like flood control, hydropower generation, and water supply Benchmark functions such as DTLZ and WFG series were often used in comparative studies to assess algorithms’ performance, as in [18,19,20,21], they consider forward and easy to solve versus real-world problems [22]. There are two types of EA parameter-setting problems categorised as (a) parameter tuning and (b) parameter
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
