基于FPGA加速的煤矿瓦斯含量预测系统

Prediction System of Coal Mine Gas Content Based on FPGA Acceleration

为了实现对煤层未开采区域瓦斯含量的快速、准确预测,以某煤矿5#煤层为研究对象,根据文献资料分析得出,煤层瓦斯含量主要受到煤层厚度、深度、顶板和底板岩性、煤层变质程度和地质结构6种因素影响。使用BP神经网络构建了瓦斯含量预测模型,并将模型移植入FPGA中,结合实际数据进行测试,结果表明,瓦斯含量预测的准确率在94%左右,预测时间为125 ms。

In order to realize the rapid and accurate prediction of the gas content in the untapped area of the coal seam,the No.5 coal seam of a coal mine is taken as the research object.According to the analysis of the literature,it is concluded that the gas content of the coal seam is mainly affected by six factors:the thickness and depth of the coal seam,the lithology of the roof and floor,the metamorphic degree of the coal seam and the geological structure.BP neural network is used to build a gas content prediction model,and the model is transplanted into FPGA,and tested with actual data.The results show that the accuracy of gas content prediction is about 94%,and the prediction time is 125 ms.