摘要
从煤层的生气潜能和储气能力两个方面分析影响煤层含气量的主控因素,认为煤层的储气能力是造成现今煤层含气量差异分布的主要因素;结合影响工区煤层含气量差异分布的主要地质因素分析,以现有的钻测资料为依据,选取相应的参数,建立适当的BP神经网络模型,对工区的煤层气含量进行预测分析.预测结果与实测资料对比分析表明:预测的煤层气含量与实测的煤层气含量之间的误差较小,且明显优于线性回归预测的结果.
Analyzed the generating and storing capacity of coal rock, the latter is the main controlling factor that affects the current distribution of coalbed gas content. Based on the field survey data Of about the gas content in coal seams in the mining areas and the analysis of the main geologic factors controlling gas content, the BP network model of prediction about gas content was established by means of BP network method of ANN ( Artificial Neural Network) theory. The model was used to predict the coalbed gas content. Compared the result predicted by BP artificial neural network with those analyzed by sample testing and predicted by regression analysis, the error of BP network method is smaller than that of regression analysis.
出处
《煤炭学报》
EI
CAS
CSCD
北大核心
2005年第6期726-729,共4页
Journal of China Coal Society
基金
国家重点基础研究发展规划基金资助项目(2002CB11702)
关键词
煤层气含量
主控因素
预测模型
BP神经网络
gas content
main controlling factors
prediction model
BP artificial neural network