Research On Reservoir Optimal Operation Nesting Considering Runoff Prediction Error

Under the background of shortage of water resource and flood disasters,to give full play to the reservoir project in water resources allocation,adjusting and flood has been the important role of,this article revolves around the rain flood resource in karst area in the view of the optimal operation of reservoirs in a river basin,with Chengbihe reservoir as a project example,the main work and achievements are as follows:(1)to Chengbihe reservoir head(1963 ~ 2014)based on the rainfall data of 52 years data,using the conservative group legal estimate flood and the flood season of design flood frequency relations,confirmed after flood standard is 2.5 times of design standards,therefore to Chengbihe reservoir in 1000 in a line on the basis of design flood process,through the same times than amplification method to derive after flood season staging in 2500 by a flood,as the incoming flood for flood regulating calculation.After determining the flood stage in the postflood season,the flood limit level of Chengbihe reservoir in the post-flood season is calculated by using the fuzzy set analysis method,and 185.5m is taken as the starting water level to perform the flood control calculus.The results show that the maximum water level in front of the dam is 186.49 m lower than the designed flood level of Chengbihe Reservoir 188.17 m in the later flood season,and the maximum downstream discharge is 2354.27 m /s lower than the designed maximum downstream discharge of Chengbihe Reservoir 3580 m /s.Therefore,the flood limit level in the later flood season is adjusted from 185 m to 185.5m,which is reliable and safe.(2)Based on the 52-year inflow runoff from the first station of Chengbihe Reservoir dam as the basic data,the long-term optimal operation and medium-and long-term optimal operation of the reservoir are nested,and a multi-objective optimal operation nested model of the reservoir is established with flood control and profit objectives in mind.Among them,the multi-objective problem is converted into a single-objective problem to solve the maximum power generation after the flood control target is converted into a hard constraint by the constraint method.In terms of solving algorithms,considering the different characteristics of algorithms and appropriate models,it is advisable to adopt deterministic algorithm for longterm optimal scheduling and intelligent algorithm for long-term optimal scheduling.Therefore,after comprehensive comparison of each algorithm,it is determined that the long-term optimal scheduling should be solved by dynamic programming algorithm,while the long-term optimal scheduling should be solved by genetic algorithm.In order to verify the optimization effect of the nested optimal scheduling model,the hydrological frequency analysis method was used to determine the three typical years of Feng(2012),Ping(1987)and Dry(1985).The nested optimal scheduling model was used to calculate the optimal scheduling decision based on the input of incoming runoff in typical years,and then the simulated scheduling was carried out.The results show that the total power generation in 2012 is 179,112,500 kW·h,the total power generation in 1987 is 206,558,300 kW·h,and the total power generation in 1985 is 158,412,200 kW·h.Compared with the electricity generation generated by long-term optimal dispatching only and actual dispatching,the annual total power generation of the reservoir has been improved.Therefore,long-term and long-term operation nesting can effectively improve the reservoir benefits.(3)the monte carlo method is used to generate the average error rate is respectively 5%,10%,15% of the error sequence,on the basis of the actual incoming runoff from 2005 to 2014 disturbance produced from different error prediction of runoff,then predict runoff input optimization scheduling nested model 300 kinds of scheduling scheme,based on the simulated scheduling scheduling scheme.In simulated dispatching,the actual inflow is controlled according to the water level process,flow rate process and output process of 300 scheduling schemes respectively,and then the advantages and disadvantages of the water level control mode,flow rate control mode and output control mode applied to Chengbihe reservoir are evaluated according to the three indexes of power generation benefit,water level overlimit rate and water abandonment.Scheduling can be simulated and the results show that in 300,the output control mode of abandoned water for an average of 392.695 million m after less than when the water level control model of ditch water,and power generation benefits than 1.33613 million kW · h flow control model,and the water level of the limit of the rate is far lower than the flow control model,from the perspective of the economic benefit and risk equilibrium output control mode for runtime Chengbihe reservoir for the optimal control model.(4)based on the output control mode analysis the influence of different error rate of Chengbihe reservoir generation benefit,drawing an average error rate of 5%,10% and 15%when years with average output error series number increasing trend of variation,and the statistical error rate under different years of average output three characteristics: mean,maximum and minimum values.Results show that with the increase of error rate series,different error rate of the average output tend to be stable for many years,and many years average output presented along with the increase of forecast error rate drop characteristics,and from the Angle of quantitative analysis shows that for many years the average of the average output,the maximum and minimum value is presented and decreased with increment of error rule,so the prediction error will affect the power generation benefit of the reservoir,and with the increment of error rate,reservoir generation benefit would be reduced.