Optimal operation model of ecological flow of hydropower station based on genetic algorithm and neural network
Yong Luo
DOI: https://doi.org/10.1016/j.heliyon.2024.e39441
IF: 3.776
2024-10-16
Heliyon
Abstract:Amid diverse water demands, the conflict between ecological flow adjustments and power generation flow adjustments in hydropower station water resource dispatching, coupled with an increasing diversity of constraints, complicates the issue of water resource optimization and dispatching. How to scientifically develop feedback regulation mechanisms to promote a virtuous ecological cycle, fine tune existing hydropower station scheduling plans, and achieve multi-objective optimization scheduling is the core issue in researching the comprehensive utilization of water resources. This study establishes a reservoir flow regulation model with the objective functions of maximizing power generation and achieving optimal average ecological protection. The model, constrained by water balance, reservoir capacity, unit output, and unit overflow, was optimized and solved using the genetic algorithm-backpropagation neural network (GA-BPNN) approach. This study establishes a reservoir flow regulation model with the GA-BPNN algorithm was applied in multi-objective optimization to identify an optimal solution that maximizes overall system performance while adhering to ecological flow constraints. The results indicate that the optimized plan has performed effectively in hydropower station operations. It not only satisfies ecological flow requirements, but also significantly enhances power generation efficiency. Specifically, power generation increased by 32,680 kw·h, a 23.85 % growth rate. The optimized plan significantly enhances the comprehensive utilization efficiency of water resources. It offers a scientific basis for the rational planning, dispatching and utilization of water resources in hydropower stations and reservoirs.