Application of the IGWO Algorithm for Flood Control in Reservoir Groups in Optimal Operation

Abstract

Abstract The operation of reservoir flood control operation is a multifaceted engineering issue characterized by complexity, several stages, nonlinearity, and many dimensions. It involves various intricate constraints and interrelated decision variables. Traditional algorithms tend to be slow and prone to local optima when solving optimization problems for flood control in reservoir groups. In recent years, with the introduction of various optimization technologies, more intelligent algorithms have been applied to optimize reservoir flood control scheduling problems in recent years. However, this remains a challenging task for large-scale reservoir group optimization scheduling problems. This work utilizes an Improved Grey Wolf Optimisation algorithm (IGWO) that incorporates Levy fly and random walk techniques for more effective optimization and scheduling of reservoir groups. Taking the Xiaolangdi Reservoir, Sanmenxia Reservoir, Luhun Reservoir, and Guxian Reservoir in Yellow River's middle and lower reaches as examples, a flood control dispatch system composed of four series and parallel reservoirs and a downstream control point at Huayuankou is studied as an example. We have established a flood control optimization scheduling model based on the Huayuankou control object, which maximizes the reduction of peak flow, and compared and analyzed the optimization results of the Improved Grey Wolf Algorithm (IGWO), Grey Wolf Algorithm (GWO), and Particle Swarm Optimization Algorithm (PSO). The results show that the Improved Grey Wolf Optimization algorithm achieves the best performance in calculating the maximum peak flow rate at Huayukou, with a peak flow rate of 18,681.1 m3/s and a peak reduction rate of 50.68%. This research offers novel perspectives and methodologies for addressing the optimization scheduling of reservoir clusters in flood control operations.