Study On On-line Identification Model Of Source Of Mine Water Inrush

Coal is the main energy in China,and it is deeply affected by the water inrush in the process of coal production.It is of great significance to identify the source of water inrush timely and accurately after the occurrence of water inrush.The most essential feature of water source identification is the analysis and processing of the water chemistry data,but the real-time acquisition of water chemistry data is a difficult problem at present.Although today's science and technology in the rapid development,the existing water inrush monitoring system cannot achieve the real-time and on-line detection,which leads to cannot real-time and accurately grasp of water quality characteristic,thus cannot timely predict water inrush.Once the water inrush occurs,how to quickly and accurately identify the source of water inrush,it is of great theoretical significance for predicting water inflow and taking relevant rescue actions.In view of occurrence of water inrush accidents frequently in coal mines,combined with the development of electrochemical sensors,the electrochemical sensor is applied to on-line detection of mine water quality,and the optimization algorithm of support vector machine based on genetic algorithm is proposed and applied to the identification of mine water inrush.Firstly,the background of the mine water hazard,the significance of water source identification and the research status at home and abroad are analyzed.Secondly,the related theory of water inrush is summarized,the regional distribution characteristics of mine water damage in China and the types and characteristics of mine water damage are analyzed in detail.Then,Several commonly used mine water inrush water source identification models are introduced,and the feasibility of using water chemistry data for water source identification is analyzed.The results show that it is feasible to use the water chemistry data characteristics to identify the source of water inrush.Based on genetic algorithm and support vector machine,adopted the idea of cross validation as the fitness function of genetic algorithm,a support vector machine parameter optimization algorithm is proposed and applied to the identification of water inrush,and the accuracy of the algorithm is compared and verified.Finally,the mine water pressure and the surrounding rock stress monitoring system are designed,and the mine water inrush model is set up.The identification of the water sourcesincluding special ion and a variety of mixed ion is analyzed and three kinds of water quality are formulated,the proposed algorithm is verified by using MATLAB software to identify the source of water inrush.The simulation results show that the proposed algorithm has high accuracy in identify the source of water inrush.Compared with the traditional water source identification algorithm,the improved algorithm has a higher recognition rate and can be used to identify the source of water inrush.