Research On Intelligent Drainage System Of Mine Based On Water Inflow Prediction

In the trend of digital transformation,safety,intelligence,green,and efficiency have become the development goals of modern coal mining enterprises.Among which underground drainage system is an important part of intelligent transformation.In response to the problems of low reliability,high energy costs and lagging automation of conventional drainage systems,this thesis proposes a short-term water inflow prediction model and an improved “peak to valley” control strategy.Besides,it establishes a pump scheduling model and designs a set of mine intelligent drainage system.The main research contents are as follows:(1)Research on mine water inflow is generally based on long-term prediction,which cannot provide early warning for water inrush accidents.This thesis proposes an improved sparrow habitus state search algorithm that improves the population initialization distribution and discovery ratio of the original sparrow algorithm and introduces a Cauchy-Gaussian variation mechanism.It proposes an enhanced state sparrow search algorithm to optimize the support vector regression of the short-term water inflow prediction model,through the weather,day type and the previous day twenty-four hours water inflow data to predict the inflow of the next day.The simulation shows that the prediction curve based on the enhanced state sparrow search algorithm optimizes support vector regression is closer to the actual trend of water inflow than the prediction results of the support vector regression(2)Aiming at the problems of single conditions,low operating efficiency and frequent start and stop of pumps of traditional control strategies,this thesis designs a "peak to valley" control strategy that subdivides the water level logic.Furthermore,by introducing water level change rate and adjusting time period dynamic adjustment control logic,an improved “peak to valley” control strategy is proposed.Experiments show that the effect of the control strategy is more stable,and the water level change trend is close to the ideal water level curve by the improved strategy.(3)In response to the problems such as power consumption pressure and high energy consumption during peak periods,this thesis builds a pump scheduling model and in conjunction with the improved “peak to valley” control strategy and surge prediction,to select the best pump scheduling scheme using Dijkstra’s algorithm to schedule pump start and stop throughout the day so that the dewatering pressure is shifted from peak times to other times.Experimental simulation shows that compared with the “peak to valley” control strategy,the pump scheduling model proposed in this thesis reduces the electricity cost by 19.3%.(4)Based on the actual scene of Jincheng Libi Coal Mine,this thesis develops a set of energy-saving and safe intelligent drainage system for coal mines,including the design of PLC control unit,data acquisition unit,intelligent video monitoring unit,etc.The multi-mode operation and water pump scheduling testing were conducted in the No.3 coal mine of Libi Coal Mine.The thesis has 42 diagrams,20 tables,and 67 references.