A Novel Hybrid Algorithm Based on Beluga Whale Optimization and Harris Hawks Optimization for Optimizing Multi-Reservoir Operation

Reservoir operation exhibits traits of nonlinearity, numerous constraints, and nonconvexity. As the number of reservoirs, such as series and parallel reservoirs, increases, the complexity of the reservoirs also increases. This study solves the complex problem using a new hybrid algorithm (MBWOHHO) based on a modified Beluga whale optimization (BWO) with Harris hawks optimization (HHO). First, in the initialization phase, an opposition-based learning strategy (OBL) is incorporated. This strategy reconstructs the initial spatial position of the population using pairwise comparisons to obtain a higher-quality initial population. Then, a differential mechanism is devised during the global search phase. This strategy enhances global exploration capabilities by cross-combining local optimal individuals with ordinary individuals. Finally, BWO and HHO are organically integrated via a population-based mechanism. This strategy effectively maximizes the strengths of both algorithms while maintaining a balance between exploration and exploitation. Several experiments are conducted across various types and complexities of benchmark functions, including 18 classical and 14 CEC2014 functions. The results of the three cascade reservoir optimization experiments show that MBWOHHO has obvious advantages over the comparison algorithms.