Study On Application Of Genetic Algorithm In Seepage Monitoring Model Of Earth-Rockfill Dam

China has the largest number of earth-rockfill dams in the world.Because seepage has an important influence on earth-rockfill dams,it is very important to study and monitor seepage to ensure the safety of earth-rockfill dams,and the establishment of seepage monitoring model is of great significance.However,when the fitting prediction effect of the seepage monitoring model is not good,there is still much room for improvement.On the basis of referring to a large number of references at home and abroad,this paper scientifically and reasonably analyzes the relationship between environmental quantity and effect quantity of earth-rockfill dam,establishes various monitoring models to fit and predict the effect quantity,and uses the optimization function of genetic algorithm to improve the fitting prediction effect of seepage monitoring model.Combined with the horizontal project: reservoir dam seepage analysis and countermeasure research,this paper carries out the research on the application of genetic algorithm in the seepage monitoring model of earthrockfill dam,and establishes the corresponding monitoring model from two aspects of pressure pipe water level and seepage flow of earth-rockfill dam.Genetic algorithm is used to optimize the seepage monitoring model.The main research contents and results are as follows:1.In this paper,based on the prototype observation data of the horizontal earth-rockfill dam,the corresponding statistical model and BP neural network model for the water level and seepage of the pressure tube are established respectively,and the genetic statistical model and GA-BP neural network model are optimized and established by genetic algorithm.Through the test index of the prediction ability of the model,it is proved that the monitoring model optimized by genetic algorithm has higher model accuracy and prediction effect.In addition,based on the monitoring model optimized by genetic algorithm,the monitoring index of pressure pipe water level and seepage established by confidence interval method can be directly used in engineering monitoring,which has strong practical significance.2.After using genetic algorithm to optimize the statistical model and BP neural network model,it is found that genetic algorithm can improve the model accuracy and prediction ability to a certain extent,but the room for improvement is limited.If the optimized model has poor model accuracy and prediction ability,it is difficult for genetic algorithm to change this situation fundamentally.Therefore,in order to obtain the optimal monitoring model,in addition to using genetic algorithm to optimize the monitoring model,it is more important to select the monitoring model with strong model fitting ability,such as BP neural network model.