里下河腹部地区除涝排水系统优化调度研究

建立了一种适合里下河腹部地区的二层除涝排水优化调度模型。第二层确定虚拟湖泊的排湖流量,第一层确定排湖流量的方式;对模型求解方法进行了改进,第二层采用Hopfild神经网络和模拟退火算法相结合的随机神经网络方法进行求解,第一层采用动态规划的方法进行求解。与以往调度方案相比,该模型能准确及时地作出面临时段的调度运行决策,提高了汛后水资源蓄存量,节省了运行费用。应用典型年份(大涝年和中等涝水年)的实测资料验证模型的合理性,结果表明模型减灾效果较好,可供类似地区决策参考。

Study on Optimizing and Operation of Water-logging Drainage in the Central Area of Li-Xia-He River

A kind of two-layer real time optimizing and operation model of water-logging drainage,which is suitable to be used in the central area of Li-Xia-He river in Jiangsu province,is founded by combining with current advanced technology of forecast.The flux excluded from the dummy lake is decided in the second layer model,and the manner of flux excluded from the lake is decided in the first layer model.The method of solving of the model is improved,the random method of NN combining with Hopfild NN and simulation anneal calculation is employed in the second layer;the method of dynamic planning is employed in the first layer.The model can make the system at its best and acquires better project benefit and save charge of running.Through analysis and calculation of the two representative years of big inundation year and middling inundation year,the rationality and precision of the model is verified.The results show that the precision of this model is high,the effect of disaster reduction is better than routine models and it is easy enough to be carried out.So it is able to give reference to the decision-makers in similar areas.