基于新型蝙蝠算法投影寻踪模型的文山州水量分配方法

摘要

基于公平和效率原则,构建文山州水量分配指标体系和水量分配投影寻踪(PP)模型.针对PP模型最佳投影方向难以确定的不足,利用新型蝙蝠算法(NBA)搜寻PP模型最佳投影方向,构建NBA-PP水量分配模型对文山州8县(市)水量进行分配.通过5个典型测试函数对NBA算法进行仿真验证,仿真结果与基本蝙蝠算法(BA)、人工蜂群算法(ABC)、布谷鸟搜索(CS)算法和差分进化算法(DE)进行对比.结果表明:通过引入生境选择策略及自适应补偿回声多普勒效应机制的NBA算法能有效平衡全局搜索和局部开发能力,寻优效果优于DE、CS、ABC和BA算法,具有较快的收敛速度、较高的寻优精度和较好的收敛稳定性与收敛可靠性;NBA-PP模型水量分配结果较目前分类权重法分配结果更科学、客观.%Based on the principles of fairness and efficiency, an index system and a projection pursuit ( PP) model were constructed for water allocation in Wenshan Zhuang and Miao Autonomous Prefecture. The novel bat algorithm ( NBA) was integrated into the PP model to solve the problem of the difficulty in determining the optimal projection direction. The established NBA-PP model was used for water allocation in eight counties in Wenshan. Five typical testing functions were used to validate the results of the NBA algorithm. The simulation results were compared with the results from the basic bat algorithm ( BA) , the artificial bee colony algorithm ( ABC ) , the cuckoo search algorithm ( CS ) , and the differential evolution algorithm ( DE ) . The results show that the NBA algorithm can effectively balance the global search and local development abilities by introducing habitat selection strategy and the self-adaptive echo-Doppler effect mechanism. Compared with the DE, CS, ABC, and BA algorithms, the NBA algorithm performs better in optimization, and it shows a higher convergence rate, higher precision, better convergence stability, and better convergence reliability. The water allocation results using the NBA-PP model are more scientific and objective than using other classification methods.