基于改进人工鱼群算法的旱限水位优化控制

Optimal Control of Drought Limit Water Level Base on Improved Artificial Fish Swarm Algorithm

针对人工鱼群算法全局搜索和局部搜索平衡能力较弱、算法运行后期搜索盲目性加大等问题,提出了人工鱼视野和步长的动态自适应调整、年际间缺水率均衡的动态控制等改进措施,建立了求解多年调节水库旱限水位的改进人工鱼群算法。通过黄河连续枯水段1990—1999年的实例研究,应用基于改进人工鱼群算法优化龙羊峡水库旱限水位,并与实际调度进行了对比。结果表明:优化控制可以合理预留调度期末水库蓄水量,流域缺水率均控制在10%左右,实现了年际间缺水均衡。

Aiming at the problem of low convergence speed of searching the global and local extreme and more blindness searching at the later stage of algorithm operation,the dynamic adjustment method for vision and step of artificial fish and the dynamic control of annual water equilibrium were adopted in improved AFSA,the improved artificial fish swarm algorithm for drought limit water level of multi-year regulating storage reservoir was established. The continuous dry year period in the 1990 s in the Yellow River basin was taken as an example,the improved artificial fish algorithm was applying in optimal control of drought limit water level and the results were compared with the actual scheduling. The results show that the optimal control can reasonablely reserve the reservoir storage at the end of regulating stage,the water deficient ratio rate can be controlled around 10% and the inter-annual water shortage equilibrium has been achieved.