Font Size: a A A

Analysis Of The Cut-off Of The Springs In Heilongtan, Lijiang Based On The Artificial Neural Network Model

Posted on:2022-01-12Degree:MasterType:Thesis
Country:ChinaCandidate:H L LiFull Text:PDF
GTID:2510306524950319Subject:Geological Engineering
Abstract/Summary:PDF Full Text Request
Groundwater resources are an important part of the earth's water resources and an important condition for human survival and development.Groundwater is the main source of water for domestic use and industrial and agricultural production in Southwest my country.The rational development and utilization of groundwater resources is also a key factor affecting the ecological environment.The Heilongtan Springs,located in the ancient city of Lijiang,Yunnan,is the source of the domestic water for residents in the ancient city of Lijiang.With the development of the society,the amount of groundwater extraction continues to increase,and problems such as unreasonable extraction appear.The groundwater level of the Heilongtan Springs continues to show The phenomenon of decline,especially when there is little rainfall or drought,the spring group will dry up,especially in recent years due to the expansion of cities and towns,the rise of industrial and agricultural industries,and well mining has become more and more serious,resulting in more and more severe dry-up of the spring group.The increase in the number of interruptions and the increase in duration have caused certain impacts on the development of Lijiang City,destroyed the local ecological environment,brought inconvenience to local residents' water for production and life,and increased the cost of water for local residents.Therefore,how to reasonably develop,utilize,and protect the groundwater resources of springs is the focus of current water resources management.How to effectively solve the problem of dry-flow of springs and find out the reasons for dry-flow of springs is the first priority at present.task.It is necessary to study the groundwater dynamics under the influence of various water conservancy projects and human activities,and establish numerical simulation models of groundwater flow under changing conditions.It is necessary to predict the equilibrium of groundwater,which is of great significance to the development and utilization of groundwater resources.In this paper,artificial neural network method is used to analyze the relationship between local rainfall and the change of Heilongtan springs.Through field geological surveys,the geological environmental conditions,hydrogeological conditions,groundwater system compensation and discharge characteristics of the study area are ascertained,the reasons for the dry-flow of the Heilongtan springs and the dry-flow process are systematically studied,and the relevant literature information is consulted and combined with the monitoring of the past years Data,using the MATKAB artificial neural network method to carry out the analysis and research on the flow and water level of the Heilongtan springs,predict the blackout time of the Heilongtan,and provide a basis for the water supply time of the Heilongtan.Use MATLAB as the tool carrier,apply least squares method to nonlinear regression analysis of the relationship between Heilongtan springs flow and water level;apply cubic exponential smoothing method to predict the water level of Heilongtan springs;apply BP neural network prediction method and neural network fitting tool Box predicts the flow of Heilongtan springs;neural network pattern recognition is used to analyze the blackout of Heilongtan springs.Use standardization as a training sample,and use a large amount of actual sample data for testing,compare the conclusions drawn by each method with the actual results,and use flow and water level as breakthrough points to conduct cut-off analysis.The main findings are as follows:(1)The main source of replenishment for Heilongtan springs is atmospheric precipitation,which is concentrated in Jiuzihai and recharges Heilongtan through Jiuzhai runoff.(2)Through statistical analysis,it is concluded that the water balance of Lijiang has undergone great changes since 2004,which is a demarcation year for the state of water balance.Before this,the dry-flow situation in Heilongtan was heavily dependent on precipitation;after this,the dry-flow situation became more and more serious due to the breakdown of the balance.The increase in groundwater extraction is the key reason for breaking the water balance in Lijiang.Due to the increasing deep drainage,the water level of the Heilongtan Well gradually drops.If the water level is not reached,the Heilongtan will be cut off.(3)Through model simulation,it is found that the discharge of Heilongtan is positively correlated with rainfall,and continuous rainfall will increase the discharge of Heilongtan.The well water level and the flow rate are also highly correlated.When the rainfall is low,the well water level will continue to decrease.It is calculated that the rainfall in the previous month in Jiuzihai has a greater impact on the well water level,and the correlation coefficient is 0.7876,The correlation coefficient of the previous two months is 0.6466,indicating that the Jiuzihai rainfall will reach Heilongtan in 1-2 months.(4)The water level of the well in Heilongtan determines whether the dry-flow occurs.When the water level of the well is lower than a certain value,no matter how much rainfall in Jiuzihai,the groundwater level must be replenished first to reach the water level before there will be flow.(5)The artificial neural network model has achieved certain results in the prediction of Heilongtan flow and dry-flow,and the predicted value is basically in line with the actual value.This model can be used to monitor the flow and dry-flow of Heilongtan,and provide water for Heilongtan.Time basis.(6)Through comprehensive analysis and calculation,it is concluded that Heilongtan can supplement water when the flow is predicted,and the amount of water supplement is calculated according to the factors affecting the flow.
Keywords/Search Tags:Lijiang, Heilongtan springs, artificial neural network, flow forecast, dry-flow analysis
PDF Full Text Request
Related items