Comprehensively utilizing the seismic,logging,drilling and outcrop data,this research studies the characteristics of the Cambrian faults and their control on the sedimentation and reservoirs in the Ordos Basin.The res...Comprehensively utilizing the seismic,logging,drilling and outcrop data,this research studies the characteristics of the Cambrian faults and their control on the sedimentation and reservoirs in the Ordos Basin.The results show that:(1)Three groups of faults striking North-East(NE),near East-West(EW),and North-West(NW)were developed in the Cambrian.The NE and near EW faults,dominated by the normal faults,are the synsedimentary faults and the main faults of the Cambrian.(2)According to the roles of faults in tectonic units and the development scale of the faults,the Cambrian faults can be divided into three grades.The second-grade faults,large in scale,controlled the boundary of the Cambrian sags of the Ordos Basin.The third-grade faults,smaller in scale than the second-grade fault,controlled the high and low fluctuations of local structures.The fourth-grade faults,very small in scale,were adjusting faults developed inside the local tectonic units.(3)The Cambrian faults had strong control on the sedimentation and reservoir of the Cambrian.Controlled by the second-grade and the third-grade faults,the paleogeographical framework of the Cambrian presents combination characteristics of the bulge-sag macro-structures and the high-low differentiation micro-geomorphology.This paleogeographical pattern not only controlled the development of the oolitic beach facies in the Cambrian but also the distribution of high-quality reservoirs.(4)Under the control of the faults,the micro-paleogeomorphological high parts closely adjacent to the margin of the Cambrian sags are the favorable exploration areas.展开更多
传统的多模态过程故障等级评估方法对模态之间的共性特征考虑较少,导致当被评估模态故障信息不充分时,评估的准确性较低.针对此问题,首先,提出一种共性–个性深度置信网络(Common and specific deep belief network,CS-DBN),该网络充分...传统的多模态过程故障等级评估方法对模态之间的共性特征考虑较少,导致当被评估模态故障信息不充分时,评估的准确性较低.针对此问题,首先,提出一种共性–个性深度置信网络(Common and specific deep belief network,CS-DBN),该网络充分利用深度置信网络(Deep belief network,DBN)的深度分层特征提取能力,通过度量多模态数据间分布的相似性和差异性,进一步得到能够反映多模态过程共有信息的共性特征以及反映每个模态独有信息的个性特征;其次,基于CS-DBN,利用多模态过程的已知故障等级数据生成多模态共性–个性特征集,通过加权逻辑回归构建故障等级评估模型;最后,将所提方法应用于带钢热连轧生产过程的故障等级评估中.应用结果表明,随着多模态故障等级数据的增加,所提方法的评估准确率逐渐增加,当故障信息充足时,评估准确率可达98.75%;故障信息不足时,与传统方法相比,评估准确率提升近10%.展开更多
三比值法是电力变压器进行潜伏性故障诊断的有效方法之一,但该方法存在缺码问题,并且对位于比值边界附近的数据易造成误判。在对大量溶解气体分析(dissolved gas analysis,DGA)数据整理分析的基础上,发现同种故障类型的数据之间,H2、CH4...三比值法是电力变压器进行潜伏性故障诊断的有效方法之一,但该方法存在缺码问题,并且对位于比值边界附近的数据易造成误判。在对大量溶解气体分析(dissolved gas analysis,DGA)数据整理分析的基础上,发现同种故障类型的数据之间,H2、CH4、C2H6、C2H4、C2H2这5种故障气体变化折线具有较强的相似性,即对同种故障的2条数据,从一条数据到另一条数,5种故障气体之间倾向于同时增加或者同时减小,而不同故障类型的数据之间,折线容易出现相反的变化,相似性差。以此规律为基础,对已有的斜率关联度进行分析探讨,对其所能刻画的斜率区间过窄的问题进行了改进,构建了负关联度计算方法,采用该方法定量分析故障气体折线的相似性,并进行故障诊断。该方法摒除了比值法的思想,保留了DGA的全部信息,能对变压器故障进行判断识别,在一定程度上克服了三比值法的缺码问题,以及在边界附近误判的问题。该方法为DGA分析提供了新的思路。实例验证了该方法的有效性。展开更多
基金Supported by the China National Science and Technology Major Project(2016ZX05007-002).
文摘Comprehensively utilizing the seismic,logging,drilling and outcrop data,this research studies the characteristics of the Cambrian faults and their control on the sedimentation and reservoirs in the Ordos Basin.The results show that:(1)Three groups of faults striking North-East(NE),near East-West(EW),and North-West(NW)were developed in the Cambrian.The NE and near EW faults,dominated by the normal faults,are the synsedimentary faults and the main faults of the Cambrian.(2)According to the roles of faults in tectonic units and the development scale of the faults,the Cambrian faults can be divided into three grades.The second-grade faults,large in scale,controlled the boundary of the Cambrian sags of the Ordos Basin.The third-grade faults,smaller in scale than the second-grade fault,controlled the high and low fluctuations of local structures.The fourth-grade faults,very small in scale,were adjusting faults developed inside the local tectonic units.(3)The Cambrian faults had strong control on the sedimentation and reservoir of the Cambrian.Controlled by the second-grade and the third-grade faults,the paleogeographical framework of the Cambrian presents combination characteristics of the bulge-sag macro-structures and the high-low differentiation micro-geomorphology.This paleogeographical pattern not only controlled the development of the oolitic beach facies in the Cambrian but also the distribution of high-quality reservoirs.(4)Under the control of the faults,the micro-paleogeomorphological high parts closely adjacent to the margin of the Cambrian sags are the favorable exploration areas.
文摘传统的多模态过程故障等级评估方法对模态之间的共性特征考虑较少,导致当被评估模态故障信息不充分时,评估的准确性较低.针对此问题,首先,提出一种共性–个性深度置信网络(Common and specific deep belief network,CS-DBN),该网络充分利用深度置信网络(Deep belief network,DBN)的深度分层特征提取能力,通过度量多模态数据间分布的相似性和差异性,进一步得到能够反映多模态过程共有信息的共性特征以及反映每个模态独有信息的个性特征;其次,基于CS-DBN,利用多模态过程的已知故障等级数据生成多模态共性–个性特征集,通过加权逻辑回归构建故障等级评估模型;最后,将所提方法应用于带钢热连轧生产过程的故障等级评估中.应用结果表明,随着多模态故障等级数据的增加,所提方法的评估准确率逐渐增加,当故障信息充足时,评估准确率可达98.75%;故障信息不足时,与传统方法相比,评估准确率提升近10%.
文摘三比值法是电力变压器进行潜伏性故障诊断的有效方法之一,但该方法存在缺码问题,并且对位于比值边界附近的数据易造成误判。在对大量溶解气体分析(dissolved gas analysis,DGA)数据整理分析的基础上,发现同种故障类型的数据之间,H2、CH4、C2H6、C2H4、C2H2这5种故障气体变化折线具有较强的相似性,即对同种故障的2条数据,从一条数据到另一条数,5种故障气体之间倾向于同时增加或者同时减小,而不同故障类型的数据之间,折线容易出现相反的变化,相似性差。以此规律为基础,对已有的斜率关联度进行分析探讨,对其所能刻画的斜率区间过窄的问题进行了改进,构建了负关联度计算方法,采用该方法定量分析故障气体折线的相似性,并进行故障诊断。该方法摒除了比值法的思想,保留了DGA的全部信息,能对变压器故障进行判断识别,在一定程度上克服了三比值法的缺码问题,以及在边界附近误判的问题。该方法为DGA分析提供了新的思路。实例验证了该方法的有效性。