Saline aquifers are the most popular waste and CO_(2)injection and storage reservoirs worldwide.This project proposes that several optimal injection positions should be investigated as hydraulic pressure-focused posit...Saline aquifers are the most popular waste and CO_(2)injection and storage reservoirs worldwide.This project proposes that several optimal injection positions should be investigated as hydraulic pressure-focused positions,in order to relieve the high demands of pump performance.The comprehensive indices(F_(i))representing the injectivity of different burial depths were obtained by using information entropy,based on the mercury injection experimental data of 13 rock samples.The results demonstrated that the burial depths of No.4,No.1 and No.2 in the Liujiagou Formation were the most suitable positions for hydraulic focused injection,which means the upper 30 m thickness could be regarded as the hydraulic focused range in the saline aquifer with an average thickness of 400 m.In addition,some laboratory experiments and in situ tests were carried out for the purpose of certifying and analyzing results,including SEM,XRD,brittleness index and logging.The results suggested that the rock samples at the No.4,No.1 and No.2 burial depth ranges have loose microstructure,weak cementation,as well as dual pores and fractures.The lithology is mainly quartz and feldspar,but the clay mineral content is high(10%-25%),which is positive for dissolution.The lithology is suitable for hydraulic fracturing to form extended cracks and micro-fissures during high-TDS(total dissolved solids)mine water injection,because of the high brittleness index.Finally,a theoretical and technical framework for high-TDS mine water injection was established,based on operating pilot engineering.Some theoretical defects and drawbacks learned from the field practices were summarized and solutions proposed.The research in this study could provide guidance and a paradigm for the inexpensive treatment of high-TDS mine water by injection and storage.展开更多
The paper analyses the principal and the methods of the optimal design of refrigerating plants.It compares the method of the static characteristics with that of the thermo dynamics,and puts the stress on the discussio...The paper analyses the principal and the methods of the optimal design of refrigerating plants.It compares the method of the static characteristics with that of the thermo dynamics,and puts the stress on the discussion of the latter which is general researched and applied in the refrigeration field.展开更多
为了解决传统机器学习对电能质量分类时人工选择特征困难导致分类效果不好的问题,本文提出了一种高度比较时间序列分析(highly comparative time series analysis,HCTSA)结合BP神经网络的分类方法,该方法既可以提取大量特征又可以筛选特...为了解决传统机器学习对电能质量分类时人工选择特征困难导致分类效果不好的问题,本文提出了一种高度比较时间序列分析(highly comparative time series analysis,HCTSA)结合BP神经网络的分类方法,该方法既可以提取大量特征又可以筛选特征,不需要人工选择特征指标。首先,使用大量算法数据库从时间序列中提取数千个可解释特征;然后,再利用线性分类器进行前向特征选择,选择出类结构中信息量最大的特征,即为最优特征。最后,用最优特征重新对扰动样本提取特征,并用BP神经网络分类。仿真结果表明,HCTSA-BP法的分类准确率达到了97.3%,比传统的小波-BP法高了8.7%。展开更多
基金supported by the National Key Research and Development Program of China(No.2023YFC3012103 and No.2019YFC1805400)the National Science Foundation of Jiangsu Province,China(No.BK20210524)+1 种基金the National Natural Science Foundation of China(No.42202268 and No.42172272)the Fundamental Research Funds for the Central Universities,China(No.2020ZDPY0201)。
文摘Saline aquifers are the most popular waste and CO_(2)injection and storage reservoirs worldwide.This project proposes that several optimal injection positions should be investigated as hydraulic pressure-focused positions,in order to relieve the high demands of pump performance.The comprehensive indices(F_(i))representing the injectivity of different burial depths were obtained by using information entropy,based on the mercury injection experimental data of 13 rock samples.The results demonstrated that the burial depths of No.4,No.1 and No.2 in the Liujiagou Formation were the most suitable positions for hydraulic focused injection,which means the upper 30 m thickness could be regarded as the hydraulic focused range in the saline aquifer with an average thickness of 400 m.In addition,some laboratory experiments and in situ tests were carried out for the purpose of certifying and analyzing results,including SEM,XRD,brittleness index and logging.The results suggested that the rock samples at the No.4,No.1 and No.2 burial depth ranges have loose microstructure,weak cementation,as well as dual pores and fractures.The lithology is mainly quartz and feldspar,but the clay mineral content is high(10%-25%),which is positive for dissolution.The lithology is suitable for hydraulic fracturing to form extended cracks and micro-fissures during high-TDS(total dissolved solids)mine water injection,because of the high brittleness index.Finally,a theoretical and technical framework for high-TDS mine water injection was established,based on operating pilot engineering.Some theoretical defects and drawbacks learned from the field practices were summarized and solutions proposed.The research in this study could provide guidance and a paradigm for the inexpensive treatment of high-TDS mine water by injection and storage.
文摘The paper analyses the principal and the methods of the optimal design of refrigerating plants.It compares the method of the static characteristics with that of the thermo dynamics,and puts the stress on the discussion of the latter which is general researched and applied in the refrigeration field.
文摘为了解决传统机器学习对电能质量分类时人工选择特征困难导致分类效果不好的问题,本文提出了一种高度比较时间序列分析(highly comparative time series analysis,HCTSA)结合BP神经网络的分类方法,该方法既可以提取大量特征又可以筛选特征,不需要人工选择特征指标。首先,使用大量算法数据库从时间序列中提取数千个可解释特征;然后,再利用线性分类器进行前向特征选择,选择出类结构中信息量最大的特征,即为最优特征。最后,用最优特征重新对扰动样本提取特征,并用BP神经网络分类。仿真结果表明,HCTSA-BP法的分类准确率达到了97.3%,比传统的小波-BP法高了8.7%。
