| Determining the source of sudden water is the primary priority of water damage prevention after mine water inrush.Only when the water source is clear can the maximum water inflow be predicted,and then corresponding water disaster prevention measures and resource utilization can be taken.Most of the wastewater generated in the mining area is directly discharged without treatment,which not only wastes water resources but also damages the ecological environment.Under this background,this thesis starts from the two perspectives of water inrush water source identification and mine water resource utilization,establishes an identification model,develops new nanomaterials,and maximizes resource utilization of water resources in mining areas.The main research contents and conclusions of this study are as follows:1)Fisher’s discriminant method based on principal component analysis to classify the hydrogeological types of national minesSelecting the national mining areas as the research object,constructing the Fisher discrimination method based on principal component analysis,and deeply analyzing the mine hydrogeological types.The contribution weight of geological types is divided into three categories,which are consistent with the actual situation.It has important reference significance for judging the hydrogeological types of mines in the country.2)Construction of DSSA-BP neural network water inrush water source discrimination modelBased on the traditional sparrow search algorithm,dynamic weights are introduced,the reverse learning strategy and differential mutation are integrated,the sparrow search algorithm is optimized,and the DSSA-BP neural network water source identification model is constructed.The results show that the model exhibited the characteristics of fast convergence speed,high accuracy and stability.3)Analysis of filling water source and chemical characteristics of groundwater in typical Shendong mining areaBy analyzing the water filling conditions of the typical Shendong mining area,the geological conditions and hydrogeological conditions of the typical Shendong mining area were identified,and the water filling water source of the mining area was deeply analyzed.,the comprehensive influence of factors such as the development height of the water-conducting fracture zone formed by the mining failure,the supply of precipitation and surface runoff,the mining method of the mine and the communication degree of the water-filled channel.4)Construction of a water inrush source identification model based on spectral line analysisThe neural network model based on spectral line analysis realizes the accurate identification of single water samples and mixed water samples in typical Shendong mining area.This method avoids the detection and analysis of complex material components and contents in water,and solves the problem of large amount of spectral data and redundant information.It has achieved the expected effect of realizing the identification of mine water inrush water source.Compared with other spectral recognition algorithms,the convolutional neural network does not need to perform feature extraction in advance,but uses the convolution kernel in the convolution layer to perform convolution operations on the spectral data to extract different features of the spectral data,which are compressed by the pooling layer.Reduce the amount of data and parameters,and then further extract and classify through the fully connected layer to reduce the amount of data operations.5)Develop new nanomaterials to realize the utilization of water resources in mining areasThe novel nanomaterials such as nano-titanium dioxide were synthesized and are used for the catalytic degradation and adsorption removal of pollutants in mining water.For example,nano-titanium dioxide achieves efficient removal of methylene blue from mining water through adsorption-catalysis synergy,and nano-titanium dioxide nanoparticles exhibit higher catalytic activity than nano-titanium dioxide nanotubes. |
PDF Full Text Request