Study On Regulation Of Reservoir Flood Input And Optimal Dispatch Under Changing Conditions

Flood disaster is one of the important problems that hinder the development of human society.It can have a significant impact on people’s production and life.In recent years,changes in underlying surface conditions and river runoff have occurred due to both climate change and human activities.It have major impact on flood control and disaster mitigation in the drainage basin.New challenges about scientific scheduling of reservoirs are presented.This paper takes the Wan’an part of Ganjiang River basin in Jiangxi Province as the research object.A numerical model of the basin was established,and the API and HEC-HMS flood forecasting modeling is carried out.The 20 fields of flood data with good representation after the year of hydrological series change were selected to rate and test the parameters of the two models,and their applicability was determined and error analysis was performed.At last,the HEC-HMS model forecast results are used for the optimized operation for reservoir.Conclusions are as follows:(1)Flooding occurs mainly in March to August,accounting for 84.9% of the total number of occurrences.Flood peak is mainly concentrated in the range of 3000～5000m~3/s,accounting for more than half of the total.Flood peak and flood peak interval time greater than7 days accounted for 95.96% of the total ratio,the percentage of the interval between rising water section is located in 2～7 days is 82.54%,the percentage of the end interval is greater than 7 days is 94.94%,and few floods occurred after September.The consistency test determined that the annual maximum flood flow series changed around 2004 and that the dominant factor was human activity.(2)(2)The continuous API model was established based on the long series of rainfall flood data in the study area.The rainfal-runoff correlation curve was stippled and runoff scheme was established.The beeline method is used to segment the flood process line.It shows that the continuous API model simulates well in calibration and validation periods,and achieves Class B accuracy.(3)With the help of ArcGIS platform for DEM data pre-processing,the study area was divided into 53 sub-basins and the area weights of rainfall stations were calculated.A spatial distribution database was established,and the construction of a digital watershed was completed.The HEC-HMS flood forecasting model was established using the digital basin.Same field floods as the continuous API model are selected for rating and validating of parameters,and the results show that HEC-HMS model outperforms the continuous API model in terms of overall simulation accuracy.(4)A single-field flood control operating model was established and optimized using an artificial bee colony algorithm.Peak-clipping rates of optimized,actual and simulated scheduling were 0.3、0.29 and 0.01 respectively,and flood water resource utilization rate was0.9 、 0.85,and 0.48.It shows that the actual dispatching work of reservoirs is better at present,and the dispatching rules currently developed lack hierarchical dispatching for small and medium floods,and operation under such rules will lead to a serious waste of flood resources.