Research On Calculation Method Of Free Outflow Of Flat Check Gate

Gate dams are an important tool to assist human development and management of water resources.China's dam construction has achieved outstanding results,and most projects have played an active role in flood prevention and drainage,irrigation and water supply,navigation and breeding,and ecological protection.Taking the Huaihe River Basin as an example,due to the needs of flood prevention and drought resistance and economic development,more than 5,400 large and medium-sized reservoirs and more than 4,200 sluices have been built in the whole river basin.However,the excessive construction of sluices and dams will inevitably cause changes in river hydrological conditions and water environment conditions,and cause consequences such as uneven water resources allocation and ecological and environmental degradation.Huaidian Sluice is a key control project on the tributaries of the Huaihe River.Accurate calculation of discharge can not only effectively reduce the losses caused by floods and droughts,but also improve the water environment and restore the ecosystem.It is of great significance to the scientific dispatching of gate and dam engineering.This paper takes the free flow of the Huaidian shallow hole sluice as an example.Based on the monitoring data of the Huaidian sluice management office from2005 to 2012,the free flow of Huaidian shallow hole sluice was simulated by different methods.The purpose is to find the calculation accuracy and differences between different methods.The main research contents of the paper are as follows:The regression equation was used to determine the flow coefficient calculation equation based on the traditional empirical formula.The obtained equation was calibrated and verified,and the flow was calculated in combination with the traditional hydraulic formula.When the obtained equation was verified,it was found that the third-order polynomial calculated well in the entire free outflow interval,with an average absolute percentage error of 0.09.The flow coefficient calculation equation of the improved piecewise function form has an average absolute percentage error of 0.075,0.068,and 0.026 when the relative opening range of the gate is 0.01 to 0.25,025 to0.45,and 0.45 to 0.60,respectively,and the simulation accuracy is significantly improved.Based on the results of different gate openings and different inflow periods,it is found that the piecewise function obtained by the research has a good application effect in the calculation of the free outflow of Huaidian sluice.MLP and RBF neural network prediction models based on different combinations of input variables are established,and MLP neural networks with better prediction results are optimized using genetic algorithms.The main factors affecting the results of MLP networks are the correlation between input variables and output variables.The increase in input variables and the number of hidden layer units can increase the accuracy of prediction to a certain extent.The RBF neural network has a lot of small-flow data in the sample,and the center of the hidden layer basis function is selected from the input sample.In most cases,it is difficult to reflect the true input-output relationship,and it is not suitable for data prediction when there are multiple input variables.After optimizing the MLP network of five input variables using GA algorithm,it was found that the improvement effect of GA algorithm is obvious,the average absolute error is reduced to 7.69m~3/s,the average absolute percentage error is reduced to 0.11,and the prediction accuracy is high.The MIKE21 model is used to calculate the free outflow of Huaidian sluice.It is found in the simulation results of the model that the MIKE model has a better simulation effect for smoothly changing data,and the model simulation has continuity,the calculation result of the previous step will affect the next step.The model has poor simulation results for small flows.When the discharge is small,the wind speed,velocity,channel roughness settings,and different incoming water conditions in the upstream may have a greater impact on it.From the model itself,these factors will be taken into account in the calculation,which will cause a large calculation error when the small flow is discharged.However,the MIKE model is very intuitive for simulating water levels and flow fields.The prediction accuracy is within a reasonable range,and it can still be used as a conventional method for predicting free flow.