摘要
为提高煤层瓦斯含量测定中损失瓦斯量补偿计算的准确度,以陕西韩城某矿正压逆流所取10组代表性煤样为研究对象,选取5种常用瓦斯损失量补偿计算模型及提出的三次方模型,在不同地点煤样、不同解吸时长、不同煤层瓦斯压力和不同解吸压力条件下试验,以拟合相关系数R^(2)、均方误差、平均绝对误差、损失瓦斯量推算误差为评判指标,模拟结果表明:解吸压力在0.6MPa以下,对数法拟合相对最小模型稳定性及精准度最高,损失瓦斯量推算相对误差、绝对误差最大值分别为-10.78%、22.23,相对最小,可选为正压逆流损失瓦斯量最佳补偿计算模型;同理,在0.6MPa以上,三次方模型稳定性及精准度最高,可选为最佳损失瓦斯量补偿计算模型。
In order to improve the calculation accuracy of gas loss compensation in the determination of coal seam gas content,taking 10groups of representative coal samples taken from a coal mine in Hancheng,Shaanxi Province as the research object,five common calculation models for gas loss compensation and the proposed cubic model were selected.Under different coal samples at different locations,different desorption duration,different coal seam gas pressures and different desorption pressure tests,the correlation coefficient R^(2),MSE(Mean Square Error),MAE(Mean Absolute Error),and the gas loss amount were taken as the evaluation indexes.The results show that when the desorption pressure is below 0.6MPa,the logarithmic method is fitted with the relative minimum model with the highest stability and accuracy,and the maximum relative error and absolute error of gas loss calculation are-10.78%and 22.23respectively,which can be selected as the best compensation calculation model for gas loss under positive pressure and countercurrent.Similarly,when the desorption pressure is above 0.6MPa,the cubic model has the highest stability and accuracy and can be selected as the best gas loss compensation calculation model.
作者
陈学习
陈江龙
黄晶晶
毕瑞卿
孙际宏
CHEN Xuexi;CHEN Jianglong;HUANG Jingjing;BI Ruiqing;SUN Jihong(School of Safety Supervision,North China Institute of Science and Technology,Langfang,Hebei 065201,China;Hebei Tangshan Petroleum Company of Sinopec Sales Co.,Ltd,Tangshan,Hebei 063000,China)
出处
《矿业研究与开发》
CAS
北大核心
2021年第12期82-87,共6页
Mining Research and Development
基金
国家自然科学基金资助项目(52174181)
中央高校基本科研业务费资助项目(3142020003)
廊坊市科学技术研究与发展计划项目(2020013162).
关键词
正压逆流取样
损失瓦斯量
三次方模型
均方误差
平均绝对误差
Positive pressure countercurrent sampling
Gas loss amount
Cubic model
Mean square error
Mean absolute error