Parameter estimation of Muskingum model using grey wolf optimizer algorithm

Area of Science:

• Hydrology
• Water Resources Management
• Computational Intelligence

Background:

• Flood routing is essential for mitigating economic and human losses from floods.
• Non-linear Muskingum models are widely used for flood routing.
• Metaheuristic algorithms offer powerful tools for optimizing complex hydrological models.

Purpose of the Study:

• To evaluate the performance of three- and four-constant parameter non-linear Muskingum models for flood routing.
• To compare the efficacy of the Grey Wolf Optimizer (GWO) algorithm against other metaheuristic algorithms in hydrological parameter estimation.
• To assess the applicability of GWO and Augmented Grey Wolf Optimizer (AGWO) for real-world flood routing scenarios, such as the Karun River.

Main Methods:

• Implementation of three- and four-constant parameter non-linear Muskingum models.
• Optimization of model parameters using the Grey Wolf Optimizer (GWO) algorithm.
• Comparative analysis with other metaheuristic algorithms including Genetic Algorithm (GA), Artificial Bee Colony (ABC), Simulated Annealing (SA), Shuffled Frog Leaping Algorithm (SFLA), and Particle Swarm Optimization (PSO).
• Validation using three benchmark examples and a real-world case study (Karun River).

Main Results:

• The GWO algorithm demonstrated superior performance in estimating hydrological parameters for the non-linear Muskingum models.
• For the three-parameter model, GWO achieved a significant reduction in Sum of Squared Quantities (SSQ) by 68% compared to the best performing alternative algorithms.
• In the four-parameter model, GWO resulted in an 18% reduction in SSQ compared to PSO, indicating its effectiveness in complex routing scenarios.

Conclusions:

• The Grey Wolf Optimizer (GWO) is a highly effective algorithm for optimizing non-linear Muskingum flood routing models.
• GWO provides more accurate hydrological parameter estimations, leading to improved flood routing predictions and reduced prediction errors.
• The study highlights the potential of GWO and AGWO for enhancing flood management strategies and mitigating flood-related risks.