Simulation And Prediction Of Mine Water Inflow Based On Time Series

Based on a comprehensive collection of Zhaojiazhai mine geological and hydrogeological data, the study found the Zhaojiazhai mine geological structure is complex, the Hutuo anticlinal structure is the main structure, with the distribution of more faults at the same time. The sandstone aquifer of the study area has Neogene gravel aquifer, the Carboniferous Taiyuan formation and Ordovician limestone aquifer. In order to strengthen the prevention and treatment of coal mine water,the water flowing fractured zone of coal seam roof and floor is studied,aquifer and mine inflow of water hydraulic connection is analyzed,the mine water inflow time series is simulated and predicted.Analysis of dynamic water level observation ledger and mine inflow ledger of Zhaojiazhai mine, found the bottom limestone aquifer of coal seams and mine inflow closely, Neogene aquifer on mine water supply is not obvious. Because the anticlinal axis bedrock thickness is small, water flowing fractured zone will involve Neogene during the mining in these area. According to the working face length is 180 m mostly, when the floor aquifuge thickness is less than 18.5m should pay attention to water inrush from coal floor.This paper focuses on the mine Zhaojiazhai mine inflow of water,organized and analyzed it with the form of time series. Firstly, on the basis of fractal theory, with the use of R/S analysis and fractal dimension method,the trend of water inflow was analyzed and predicted.We found that mine water inflow time series of 2009 to 2013, are with long range correlation year after year or five years overall, and the future water will keep the original trend of slow decline.Gray theory as the basis, this paper study the correlation during the mine water inflow time series and each aquifer drawdown depth. We get that, each correlation between them is greater than 0.5, and the biggest one is from L7-8. Application of different grey modeling method to simulate and predict the water inflow of mine on 2013, analog value relative error is 0.9% and 4.8% respectively, shows the grey modeling effect is very good. From the GM(1,1) model we get the mine inflow in the first quarter of 2014 are respectively 1063m3/h、1053m3/h、1044m3/h, from the GM(1,2) model we get the results are 1111m3/h、1108m3/h、1106m3/h respectively.