Simulation-Optimization Approach Based on Meta-Model in Optimal Design of Inter-Basin Water Transfer System

In this research the simulation-optimization (S-O) approach is applied to determine the optimum capacity of water transfer systems in “BehestAbad” to “Zayandehrood” Inter-basin water transfer project. In this regard, a Multi-objective Multi-Cooperative Swarm Particle Swarm Optimization (MOMCSPSO) algorithm as an optimization technique is linked to Water Evaluation and Planning system (WEAP) as simulation model. Due to the immense computational efforts of simulation-optimization approach specifically in an evolutionary algorithm application and simulation model frequent recalling requirements, intensive computational costs are expected. To alleviate this challenge, a surrogate based S-O approach is developed in this research. Artificial Neural Network (ANN) as surrogate model substituted WEAP. The objective functions of the problems are minimizing the environmental and domestic water demand shortages at the origin and destination basins separately, and minimizing the water transfer system costs. The results show that the proposed optimization algorithms are able to offer reasonable and diverse solutions that has a good utility in terms of management decisions. Furthermore the developed techniques in this research are effective on computational time reduction whereas the desired accuracy is achieved in water resources objective function estimations.