Real-time Monitoring And Dynamic Forecast Of Coal Mine Ground Water-level Based Neural Net

With the rapid development of our economics, the meet of coal is larger and larger. As our known, the coal industry is one of the most dangerous energy industries. Coal exploitation is subjected to much restrictions, and ground water is an important factor of threatening to coal security exploitation. Recently, many important accidents happened on news or reports, which jeopardized to people's lives and wealth. So the research of ground water-level real-time monitoring and dynamic forecast has important valuable engineering meanings. This paper analyses and compares the current methods of monitoring ground water-level, such as auto-record, wire monitoring, non-wire monitoring, and completes a non-people remote monitoring using GSM Modem Short Message Service technology. This system reduces labor intensity and improves work efficiency. This paper establishes non-linear model between some influent-factors and ground water-level with neural net knowledge, and makes single influent-factor experiment and multi influent-factor experiment using a lot of past data, and forecasts the ground water-level trend in the coming period, and gets better effect, which provides references for subsequence work of coal security production. This monitoring system implements data input function, real-time monitoring function, water-level forecast function, making ground water-level isoline graph with the Visual Basic 6.0 and Surfer 8.0 develop tools based GSM Modem SMS and neural net technologies expatiated in the above paragraphs, which completes real-time monitoring and forecast coming ground water-level. The result indicates that the above method implements the forecast of ground water-level and gets the ideal result, combining the ground water-level of Nantun Coal Mine, Yankuang Group, laying the foundation of further system development.