基于交会图和多元统计法的神经网络技术在火山岩识别中的应用

针对火山岩储层的特殊性(复杂性、离散性和随机性),应用BP神经网络技术对火山岩测井解释中岩性识别问题进行了研究。该方法的技术关键是样本集和初始权重的建立,以及模型的优选。本文提出了一种基于交会图和多元统计法的学习样本生成方法,即根据取心岩样的地球化学和岩石学研究,用交会图技术建立样本集,采用聚类分析和距离判别法确定初始权重。将研究方法应用在松辽盆地杏山地区火山岩岩性识别问题中,取得了很好的效果,岩性解释符合率高于90%。文中通过四种岩性识别处理模式的对比研究,表明赋权重处理模式为最优处理模型。在神经网络模型预测过程中,需充分利用已有的地质经验和测井曲线信息建立典型可靠的样本文件,同时考虑神经网络方法中各种因素的影响,优选模型和计算参数才能使预测结果符合实际情况。     

移动通信系统中功率控制研究

在阐述功率控制在移动通信系统中的发展过程的基础上，着重研究了WCDMA系统的功率控制，最后对功率控制未来的研究方向做了简要的分析。     

复杂土坝的渗流安全分析评价

根据福华山土坝的渗流观测资料,建立了测压管水位的回归统计模型,定量分析了各影响量(水位分量、时效分量)对测压管水位的影响效应。在此基础上,结合观测资料,采用有限元法反演了大坝各填筑区的渗透系数,并综合评价了大坝的渗流性态。     

博宁联合化疗治疗骨转移癌的疗效分析

目的 :分析单纯化疗与博守联合化疗对实体瘤骨转移引起疼痛治疗的效果。方法 :42例病人随机分为单纯化疗组和化疗联合博宁组 ,对疼痛缓解程度和活动能力改善等进行对照研究。结果 :化疗联合应用博宁组疼痛缓解率及肿瘤变化方面明显优于单纯联合化疗组 (P <0 .0 5 ) ,而在活动能力改善方面则无明显差别。结论 :博宁联合化疗治疗骨转移癌引起的疼痛效果优于单纯联合化疗     

射孔完井出砂预测新模型及其在射孔优化中的应用

分析了射孔完井的油层出砂机理，以线弹性理论为基础，考虑变温应力和流体渗流作用的拖曳力影响，建立了一种新的射孔完井出砂预测模型。应用该模型研究射孔方位、射孔参数对油井出砂的影响，结果表明，深穿透射孔弹比大孔半径射孔弹能更好地避免孔道出砂，采用定向射孔能有效地减少油井出砂问题。其结果对砂岩油藏射孔优化设计具有一定指导意义。     

天然气井作业屏蔽暂堵技术的研究与应用

在天然气井进行井下作业前，将暂堵剂挤入生产层位附近，以暂时封堵生产层，使地层压力不被释放，井筒内的压井液不污染地层，待施工作业结束后，及时进行解堵，确保天然气井不受伤害或少伤害，从而恢复天然气井的生产能力。针对天然气井大修作业时，经常补充压井液，由于作业时间长，大量压井液进入地层，造成严重污染，导致天然气井恢复生产困难的问题，研制出一种新型气井作业的屏蔽暂堵剂，具有污染小，易解堵等特点，有效解决了大修污染问题，现场应用取得良好效果。     

变频电子阻垢技术的工业应用

中国大唐发电有限责任公司徐塘分公司利用变频电磁阻垢技术在4号发电机组循环冷却水系统进行了成功应用.从理论上比较了物理法与化学法水处理技术的差异.工业应用结果表明,变频电磁技术不仅阻垢效果明显,而且具有安装维护简便,适应性广,运行费用低,对水体无污染等特点.可以替代化学阻垢法.     

CO2吞吐工艺操作参数的整体优化设计

针对ZY油田原油CO2吞吐的工艺过程，建立了CO2吞吐数值模拟模型。应用数值模拟方法对CO2吞吐进行了实例分析，首先对该井的CO2吞吐参数进行了单因素敏感性分析，分析了CO2周期注入量、注入速度、焖井时间、采液速度等因素对CO2吞吐效果的影响。在此基础上进一步运用正交试验方法对CO2吞吐参数进行了优化，通过计算表明用正交试验优选法得到的优化方案比单因素敏感性分析所得的优化结果更为合理。     

Range Assignment for Biconnectivity and k-Edge Connectivity in Wireless Ad Hoc Networks

Depending on whether bidirectional links or unidirectional links are used for communications, the network topology under a given range assignment is either an undirected graph referred to as the bidirectional topology, or a directed graph referred to as the unidirectional topology. The Min-Power Bidirectional (resp., Unidirectional) k-Node Connectivity problem seeks a range assignment of minimum total power subject to the constraint that the produced bidirectional (resp. unidirectional) topology is k-vertex connected. Similarly, the Min-Power Bidirectional (resp., Unidirectional) k-Edge Connectivity problem seeks a range assignment of minimum total power subject to the constraint the produced bidirectional (resp., unidirectional) topology is k-edge connected. The Min-Power Bidirectional Biconnectivity problem and the Min-Power Bidirectional Edge-Biconnectivity problem have been studied by Lloyd et al. [23]. They show that range assignment based the approximation algorithm of Khuller and Raghavachari [18], which we refer to as Algorithm KR, has an approximation ratio of at most 2(2 – 2/n)(2 + 1/n) for Min-Power Bidirectional Biconnectivity, and range assignment based on the approximation algorithm of Khuller and Vishkin [19], which we refer to as Algorithm KV, has an approximation ratio of at most 8(1 – 1/n) for Min-Power Bidirectional Edge-Biconnectivity. In this paper, we first establish the NP-hardness of Min-Power Bidirectional (Edge-) Biconnectivity. Then we show that Algorithm KR has an approximation ratio of at most 4 for both Min-Power Bidirectional Biconnectivity and Min-Power Unidirectional Biconnectivity, and Algorithm KV has an approximation ratio of at most 2k for both Min-Power Bidirectional k-Edge Connectivity and Min-Power Unidirectional k-Edge Connectivity. We also propose a new simple constant-approximation algorithm for both Min-Power Bidirectional Biconnectivity and Min-Power Unidirectional Biconnectivity. This new algorithm applies only to Euclidean instances, but is best suited for distributed implementation. A preliminary version of this work appeared in the proceedings of the 2nd International Conference on AD-HOC Network and Wireless (Adhoc-Now 2003). Research performed in part while visiting the Max-Plank-Institut fur Informatik. Gruia Calinescu is an Assistant Professor of Computer Science at the Illinois Institute of Technology since 2000. He held postdoc or visiting researcher positions at DIMACS, University of Waterloo, and Max-Plank Institut fur Informatik. Gruia has a Diploma from University of Bucharest and a Ph.D. from Georgia Insitute of Technology. His research interests are in the area of algorithms. Peng-Jun Wan has joined the Computer Science Department at Illinois Institute of Technology in 1997 and has been an Associate Professor since 2004. He received his Ph.D. in Computer Science from University of Minnesota in 1997, M.S. in Operations Research and Control Theory from Chinese Academy of Science in 1993, and B.S. in Applied Mathematics from Tsinghua University in 1990. His research interests include optical networks and wireless networks.