Research And Application On Comprehensive Flood Control Dispatching Of Reservoir Groups And Flood Detention Areas

As one of the most serious natural disasters in the world,flood impacts extensively and is easy to case extremely high losses.China,also affected by floods.For instance,for three consecutive years from 2016 to 2018,the Yangtze River Basin happened large floods in middle and lower,middle,and upper reaches to make extreme pressure for the flood control security.However,with the construction and operation of large-scale cascade reservoirs,their huge flood control capacity can heavily cut down the peak of flood and store flood.Besides,other excess flood can be disposed with flood detention areas in the middle and lower reaches,which can effectively reduce the flood control pressure of key cities along the Yangtze River and flood control points,and decrease losses to the maximum extent.Thus,a prefect flood control and dispatching system is an effective method to reduce or even avoid flood losses.Combined and integrated dispatching of cascade reservoir groups and flood detention areas in the catchment not only bring great benefits in flood control,but also exist many theoretical systems to be improved and practical engineering problems to be solved.It is mainly in the following three aspects:(1)In the aspect of reservoir group dispatching,the information sharing is not timely because the reservoir group belong to different dispatching institution.And the existing methods can still not forecast accurate value of runoff.Higher time cost will product when take full Saint-Venant equations into consideration.To propose high-efficiency coupled models between reservoir groups and river networks are the one of research hotspot.(2)With the shark development of social economy,the construction of flood detention areas or flood protected areas are gradually to be perfect.Huge property will be damaged once to diverse flood actively or passively.Hence,to study elaborate flood routing simulation techniques and to develop evacuation routes planning methods are the other key research solution.(3)Complex flood control system involve multi-divisional,multi-region,and multi-dispatching-body.How to contract each parts precisely and fast forecast flood control state has a great and important significance.In conclusion,flood control dispatching of basin with reservoir groups and flood detention areas is a solution which should be dispatched by multi-bodies independently and uniformly.New theories,methods and techniques should be developed in combination with other subjects.For this purpose,we took the Three Gorges Reservoir,the cascade reservoir groups in the upper reaches of the Yangtze River,flood protected area of main-dike in Yangtze River from Hannan to Baimiao,and Hanbei River Basin as examples.Further researches on fast and accurate simulation of river flood,reservoir group dispatching and flood detention area elaborate diversion operation,starting from establishing high-efficiency runoff simulation and prediction model,were made by using numerical analysis,what-if analysis,mathematical simulation,contrastive analysis and engineering verification.The main contents and innovative achievements of this paper are as follows:(1)To solve the simulation of complicate dendritic river network,a fast and high-precision one-dimensional model had been proposed without solving junction equation.Based on Saint-Venant equations,semi-implicit?is employed to discretize the water level gradient term of momentum equation,finite element volume method was employed to discretize equations of continuity,Eulerian-Lagrangian method was adopted to solve the advection term and finally to realized fast and high-precision simulation of dendritic river network by prediction-correction method.Three typical tests were used to verified the model reliability and feasibility.Moreover,two real river network were also tested.The results showed that the proposed model simulated well for dendritic river network,and was also useful for other river types.Compared to the results of HEC-RAS and MIKE 11,the proposed model(H1DM)was in the lead of calculation efficiency,which could provide powerful supports for solving the dispatching of cascade reservoirs coupled river network.(2)The sensitivity factors affecting the flood propagation law and the change law were explored.The concept of flat section in front of the dam had been defined and the capacity change law were calculated and analyzed.Supposing that river channels of Three Gorges happened gradual and shark flood waves under different water level of reservoirs,different peaks of flood,different surge waves,and different maximum flow durations,and found out that the most sensitive factor is water level——Under the current flood condition,the flood propagation time decrease 0.5 h while the level increase 1 m.And then,capacity change rule of the Three Gorges Reservoir with topography data on the year of 2017 was analyzed.Compared with annual average static reservoir capacity,the annual average dynamic capacity could be exceed 4%.The land zone in front of the dam was also analyzed which could initially consider that it was about 152 km far away from the dam.The results can be helpful to real-time flood control dispatching.(3)Center on the solution of the dispatching of reservoir groups in the upper reaches of the Yangtze River,the effect on flood control benefit of the Wudongde-Baihetan cascade reservoirs was analyzed.A deep-learning-based flood control model for reservoir groups which could fast forecast discharge magnitude was built.And considering flat section in front of the dam,a H1DM-driven conventional flood control dispatching model was proposed.The dispatching process with and without the Wudongde-Baihetan cascade reservoirs were calculated.The results showed that this cascade could further reduce the flood control pressure.Then,to analyze the dispatching process of the cascade reservoirs in the middle of Jinsha River basin.The proposed model could fast provide accurate forecast of discharge magnitude.In final,aiming at forecasting the inflow of the Three Gorges,a comparison between general model and the new model was made.The results indicated that the coupled model had higher precision in prediction and could provide ambient field and quantitative basis for inflow flood forecast in real-time flood control dispatching.(4)In allusion to the inhibition,guide,and storing of linear features in the process of flood routing,an elaborate modelling method was proposed.Introduced capacity constrained network,Quickest Flow Span(QFS)is defined.An Evacuation Routes Planning model was built to minimize the economic losses of flood detention areas.Capacity Constrained Quickest Flow Route Planner algorithm(CCQFRP)is proposed.This thesis took flood protected area of main-dike in Yangtze River from Hannan to Baimiao as the experimental object,analyzed flood inundation situations under different flood sources,evaluated the corresponding flood risk,and using CCRP algorithm and CCQFRP algorithm to solute routes planning problem.The results indicated that the proposed models could reflect real flood routing process effectively,and CCQFRP algorithm reduce almost 71%time cost which had engineering application value.(5)Taking the comprehensive system construction of flood control and dispatching as the background,the flood routing dynamic boundary of flood protective embankment was innovatively proposed.A group of full dynamic characteristic surface,composed of discharge,time,and inundation area,were ascertained.A spatio-temporal forecasting repository of discharge,moment,and inundation area was established by BP neural network.And a coupled basin optimizing dispatching model between sphev reservoirs and flood protective embankments was built.The comprehensive optimizing dispatching of basin was realized.The Hanbei River Basin was selected to be the research object.Historical optimizing dispatching were simulated in the year of 2007,2008 and 2016.The results showed that the optimizing dispatching has obvious effect on cutting down the peak of flood in flood control point.For Tianmen Station,the maximum peak clipping was 130 m~3/s,which was about 20.8%of historical data.Besides,the optimizing flood diversion dispatching reduce 87.83%inundation area.Finally,the information techniques to establish the integrated flood control system were discussed based on micro-service architecture,and the architecture and function of the joint flood forecasting and dispatching were introduced.