Decentralized calibration process for distributed water resources systems using the self-adaptive multi-memory melody search algorithm

Having systematic simulation and optimization models with high computational accuracy is one of the most important problems in developing decision support systems. In the present research, a specific methodology was proposed for decentralized calibration of complex water resources system models by using the structural capabilities of the melody search algorithm. This methodology was implemented in the framework of a self-adaptive simulation-optimization model that helps fine-tune complex water resources models by introducing a new definition of the way sub-memories are related to each. The introduced structure aims to achieve the highest possible level of consistency, which is estimated by using different criteria, between model results and observed data at several control points of surface flows. The introduced strategy was put to the test in developing a water resources model for the Great Karun Watershed, Iran, and was found to produce accurate results compared to some other well-known optimization algorithms such as GA, HS, PSO, SGHS, EMPSO, and SaMeS. In an attempt to determine the effect of calibration on water resources system modeling, 16 calibration models of different dimensions are developed and their computational costs are compared in terms of their computation time and effects on the accuracy of the results.

Automated summary

The research proposes a method for decentralized calibration of complex water resources system models using the melody search algorithm within a self-adaptive simulation-optimization model framework. The introduced structure aims to achieve the highest possible level of consistency between model results and observed data, and was validated in developing a water resources model for the Great Karun Watershed in Iran.