土耳其盐穴地下储气库大井眼钻井技术

土耳其Tuz G?lü盐矿区拟建造地下天然气储气库,存在2个可能性风险:矿床实际深度与原计划钻探深度的偏离;钻井与建腔过程中侵出CO2气体,因此该气库建造存在一定的地质及工程风险。为加快该项目进展,满足盐穴储气库钻井要求,通过优化方案采用国内没有应用过的生产套管大尺寸井眼井身结构,采用液压扩孔器扩眼,并形成钻井配套技术。文中从井身结构、钻头优选及螺杆钻具应用、钻井液体系优选、固井以及钻探效果等方面,阐述Tuz G?lü天然气地下储气库钻井技术。通过实施,有效降低了钻井事故复杂时率、缩短钻井周期,机械钻速得到提高,取得了显著的经济效益。     

华润矿综放工作面顶煤放出率提升措施的探究

为促进华润矿煤炭开采效率的进一步提升,对华润矿综放面顶煤放出率提升措施展开探究,从预先爆破、支架优化、放煤工艺优化、加强工作面初末采及上下端头顶煤回收四个方面作出理论分析.     

Development of mobile miniature natural gas liquefiers

With increasing consumption of natural gas (NG), small NG reservoirs, such as coalbed methane and oil field associated gas, have recently drawn significant attention. Owing to their special characteristics (e.g., scattered distribution and small output), small-scale NG liquefiers are highly required. Similarly, the mixed refrigerant cycle (MRC) is suitable for small-scale liquefaction systems due to its moderate complexity and power consumption. In consideration of the above, this paper reviews the development of mobile miniature NG liquefiers in Technical Institute of Physics and Chemistry (TIPC), China. To effectively liquefy the scattered NG and overcome the drawbacks of existing technologies, three main improvements, i.e., low-pressure MRC process driven by oil-lubricated screw compressor, compact cold box with the new designed heat exchangers, and standardized equipment manufacturing and integrated process technology have been made. The development pattern of “rapid cluster application and flexible liquefaction center” has been eventually proposed. The small-scale NG liquefier developed by TIPC has reached a minimum liquefaction power consumption of about 0.35 kW·h/Nm³. It is suitable to exploit small remote gas reserves which can also be used in boil-off gas reliquefaction and distributed peak-shaving of pipe networks.     

可调偏迈步式胶带自移机尾的研究

随着掘进工作面自动化程度的提高,掘进巷道速度也在日益加快,而现有的顺槽皮带机机尾移动缓慢且移动时工作量大,需要两三人配合操作,存在一定的安全隐患。针对以上问题,研发可调偏迈步式胶带自移机尾,具有与悬挂二运皮带转载机搭接行程长,可以迈步式前进,自动调偏等功能,同时,可以降低工人的劳动强度,加快巷道掘进速度。     

粉煤灰砖抗折强度测量结果不确定度的评定

通过对粉煤灰砖抗折强度测量结果不确定度的评定,分析出影响试验结果的各种因素,为提高检测水平和精确度提供指导。     

基于ANP-SWOT的上海公路废弃物资源化发展对策研究

随着公路系统的新陈代谢，大量废弃物给城市的资源与环境协调发展带来了巨大压力，以沥青路面再生技术为路径的公路废弃物资源化符合循环经济的理念，是我国公路建设迈向生态文明的必然选择。通过 ANP-SWOT 对上海公路废弃物资源化发展战略进行实证分析，发现当前形势下其优势和机遇较为明显，确定以 SO 策略为核心的推广战略为最佳选择，并提出一系列配套发展策略，为推动上海公路废弃物资源化提供参考。     

Effect of Biochar as Reductant on Magnetizing-roasting Behavior of Pyrite Cinder

The effect of biochar substituted for anthracite as reductant on magnetizing-roasting pyrite cinder was investigated.The key of magnetizing-roasting is the gasification reaction between reductants and CO2.Since biochar could react with CO2 more rapidly at lower temperature,the reactivity of biochar is better than that of anthracite.The gasification of biochar could produce reducing condition ofφCO/(φCO+φCO2)about 10%-20% between 700-800 ℃,which is in accord with the atmosphere and temperature of Fe2O3 reduction.So it is beneficial to the reduction of iron mineral of pyrite cinder.Compared with anthracite,biochar could decrease the roasting temperature from825 to 750 ℃ and roasting time from 20 to 15min,which shows that a better effect of magnetization could be obtained in the condition of lower temperature and shorter time.Using biochar as reductant,iron concentrate extracted from pyrite cinder as about 64%iron grade could be produced,and the recovery is over 90% under the condition of above 90% grinding particle less than 0.045 mm and magnetic intensity of 0.124-0.194 T.     

马钢张庄矿高压辊磨机装配工艺研究

针对马钢张庄矿高压辊磨机辊套疲劳断裂严重、对辊精度不能满足生产要求等问题，马钢重机公司根据张庄矿实际生产要求对对辊进行国产化，国产化内容包括辊轴、辊套等零部件的新制以及装配工艺攻关，通过技术研发同时结合实际生产经验，成功对高压辊磨机辊套与辊轴以及辊轴与轴承进行装配，完成了张庄矿高压辊磨机整对对辊的国产化，为该型设备后续生产及优化提供了理论和实践参考依据。     

基于注意力机制与文本信息的用户关系抽取

随着社交媒体的发展,用户之间的关系网络对于社交媒体的分析有很大的帮助。因此,该文主要研究用户好友关系检测。以往的关于用户好友关系抽取的研究主要基于社交媒体上的结构化信息,比如其他好友关系,用户的不同属性等。但是,很多时候用户本身并没有大量的好友信息存在,同时也不一定有很多确定的属性。因此,我们希望基于用户发表的文本信息来对用户关系进行预测。不同于以往的潜在好友推荐算法,该文提出了一种基于注意力机制以及长短时记忆网络(long short-term memory,LSTM)的好友关系预测模型,将好友之间的评论分开处理,通过分析用户之间的评论来判断是否具备一定的好友关系。该模型将好友双方信息拼接后的结果作为输入,并将注意力机制应用于LSTM的输出。实验表明,用户之间的评论对于好友关系预测确实有较大的实际意义,该文提出的模型较之于多个基准系统的效果,取得了明显的提升。在不加入任何其它非文本特征的情况下,实验结果的准确率达到了77%。