基于神经网络的矿山多源信息融合方法研究

Research on Mine Multi Source Information Fusion Method Based on Neural Network

在分析煤矿六大安全避险系统现状的基础上,针对煤矿六大系统相互独立、各类数据综合利用率低等问题,研究矿山多源数据融合的关键技术,建立了模糊神经网络模型,设计了矿山多源信息融合系统。系统采用IFIX组态软件为开发平台,实现井下设备监测、人员定位、通风供水、紧急避险等功能的互联互通,在淮北矿业集团朱仙庄煤矿对该系统进行了测试,结果表明,系统能很好地解决煤矿六大系统相互独立的问题,提高了矿井多源数据的综合利用率。

On the basis of analyzing the present situation of the six major safe-hazard systems in coal mine, the key technology for multi-source data fusion of mines was studied and the fuzzy neural network model was established, aiming at the independence of six major coal mine systems and the low comprehensive utilization of various types of data. The mine multi-source information fusion system was designed. The system adopts the IFIX configuration software with OPC interface as a development platform and supports the interoperability of downhole equipment monitoring, personnel positioning,ventilation water supply and emergency avoidance functions with mathematical models. The system was tested in Zhuxianzhuang coal mine of Huaibei mining group. The feasibility of the results confirmed that the system solved the problem of the independence of the six major coal mine systems and improved the overall utilization of data.