基于改进沙猫群算法的水库群防洪优化调度研究

Research on Flood Control Optimization and Scheduling of Reservoir Groups Based on Improved Sand Cat Swarm Algorithm

水库防洪调度对于有效减少洪水灾害、保障人民生命财产安全至关重要。此过程是个多阶段、非线性的、高纬度的工程问题,具有许多复杂的约束条件和相互依赖的决策变量。为了提高水库群优化调度问题的求解效率,充分发挥水库群协同防洪能力,提出了改进的沙猫群算法(Sand Cat Swarm Optimization Algorithm,SCSO),利用Cubic混沌映射策略实现调度方案的分散均匀性,引入鲸鱼算法的螺旋搜索策略提高种群的局部搜索和全局搜索能力,融合麻雀算法后阶段的预警机制增加算法后期全局搜索的能力,使用经典测试函数和秩和检验对算法的精度进行检验,结果表明,改进后的沙猫群算法的收敛速度和精度都得到了明显的提升;并首次将算法运用在水库群防洪优化调度上,建立防洪控制点处最大削峰准则模型,对黄河中下游5座水库联合防洪调度系统应用研究,同时,将改进的沙猫群算法(ISCSO)与原始沙猫群算法(SCSO)、蜣螂算法(DBO)的优化结果进行对比分析,其中DBO算法求得的控制点峰值流量为21274.3 m^(3)/s,削峰率为46.62%,SCSO算法求得的控制点峰值流量为21248.6 m^(3)/s,削峰率为46.68%,ISCSO算法求得的控制点峰值流量为20687.1 m^(3)/s,削峰值最率最大,为48.09%。结果表明,改进的沙猫群算法在解决水库防洪调度问题中削峰效果最好,且有效实现下游错峰效果,保证了下游河道以及防洪控制点的安全。研究成果为解决水库群防洪优化调度提供了新的思路和方法。

Reservoir flood control scheduling is crucial for effectively reducing flood disasters and ensuring the safety of lives and property.This process is a multi-stage,nonlinear,high-dimensional engineering problem with complex constraints and interdependent decision vari⁃ables.To improve the efficiency of solving the optimal scheduling problem of reservoir groups and fully utilize their coordinated flood control capabilities,this paper proposes an Improved Sand Cat Swarm Optimization Algorithm(SCSO).The algorithm incorporates the Cubic chaot⁃ic mapping strategy to enhance the dispersion and uniformity of the scheduling plan.Additionally,the whale algorithm′s spiral search strategy is introduced to improve the population′s local and global search capabilities,and the sparrow algorithm′s warning mechanism is integrated to boost global search ability in later stages.The algorithm′s accuracy is validated using classical test functions and rank sum tests.Results demonstrate significant improvements in the convergence speed and accuracy of the improved Sand Cat Swarm Optimization(ISCSO)algo⁃rithm.For the first time,this algorithm was applied to the optimization and scheduling of flood control in reservoir groups.A maximum peak shaving criterion model was established at flood control points,and the joint flood control scheduling system for five reservoirs in the middle and lower reaches of the Yellow River was studied.Comparative analyses were conducted with the original Sand Cat Swarm Optimization(SC⁃SO)and the Dung Beetle Optimization(DBO)algorithms.The peak flow rate at control points obtained by the DBO algorithm was 21274.3 m^(3)/s,with a peak shaving rate of 46.62%.The SCSO algorithm yielded a peak flow rate of 21248.6 m^(3)/s,with a peak shaving rate of 46.68%.The ISCSO algorithm achieved a peak flow rate of 20687.1 m^(3)/s,with the highest peak shaving rate of 48.09%.These results indicate that the ISCSO algorithm achieves the best peak shaving performance for reservoir flood control scheduling,effectively shifting downstream peaks and ensuring the safety of downstream rivers and flood control points.The findings of this research provide new insights and methods for opti⁃mizing flood control scheduling in reservoir groups.