地震属性神经网络油气模式识别技术及其在东海的应用

神经网络油气模式识别技术是综合利用地震属性进行油气预测的技术之一 ,它首先通过计算获得多种地震属性 ,综合分析找出对储层油气比较敏感的地震属性组合 ;然后收集油气井与非油气井的井旁道地震属性 ,组成学习样本并进行神经网络学习 ;最后利用学习结果对储层进行油气预测。该技术在东海某工区的应用结果表明 ,振幅统计类和复数道统计类地震属性是对该地区储层油气最敏感的地震属性组合。神经网络油气模式识别技术可以作为东海地区储层油气预测的一种手段。     

基于感知器的快速彩色图像无损压缩算法

针对彩色图像像素和颜色分量之间存在的相关性，将神经网络中的感知器技术引入彩色图像的无损压缩算法中，提出了一种新的预测模型。应用感知器的自学习和自适应能力，对预测值进行自适应调整，使算法在运行过程中具有很小的预测残差并具有较小的动态范围。同时应用颜色空间变换来减小颜色分量间的相关性。相对于新的国际标准JPEG—LS，这种预测模型具有较低的计算复杂性。对比实验结果说明，这种算法的性能明显优于传统算法，在压缩比损失很小的前提下，其执行速度高于JPEG—LS。     

Distributed Speech Recognition Systems and Some Key Factors Affecting It''''s Performance

1　IntroductionWiththerapiddevelopmentofInternetandmobilecommunications,theresearchonwirelessmobilenetworkshasbecomethenewhotspotinin formationtechnologyfield .Asweknow ,mobileterminals (mobilephone,forexample)arethetooltohelppeopletoaccessanyinformationtheyneedbyanymediaatanytimeandanywhere .However,itissometimesdifficultformobileterminalstoplaysucharolebecauseofthelimitationssuchaslimitedca pacity ,lowpower,andinconvenienceinoperating .So,peoplehopetofindanotherwaytoaccessnet workinformatio…     

环境因子的定义及研究现状

从普遍接受的三个基本假设出发，讨论环境因子的定义和失效机理不变的约束条件，重点综述环境因子的研究现状和常用方法，并探讨引入反应论模型解决环境因子预测问题的可能途径。     

光纤码分多址技术在城域接入网中的应用

分析了光纤码分多址技术O -CDMA在城域接入网中的应用。回顾了O -CDMA技术的发展历史 ,详细地介绍了O -CDMA技术的特点 ,其具有的公平性、灵活性、网络控制管理、对服务区处理能力以及高的安全性使O -CDMA技术成为较好的城域接入网技术。阐述了目前在应用中O -CDMA存在的一些技术方面的问题     

DCS在采暖锅炉房中的应用

介绍了Jx—300X DCS系统的组成、功能、特点以及热水锅炉主要控制方案及其在锅炉测控方面的应用。     

大规模MPLS网络中的新兴技术

随着网络规模越来越大,选择一个完善的网络管理框架并有效降低标签需求量对于MPLS网络而言显得十分重要。为此,探讨了基于策略的MPLS管理以及Trainet等近年来出现的大规模MPLS网络新兴技术,并与现有相关技术进行了比较。     

Assessment of Bonelike® graft with a resorbable matrix using an animal model

Synthetic bone grafts have been developed to provide an alternative to autografts and allografts. Bonelike® is a patented synthetic osteoconductive bone graft that mimics the mineral composition of natural bone. In the present preliminary animal studies a user-friendly version of synthetic bone graft Bonelike® have been developed by using a resorbable matrix, Floseal®, as a vehicle and raloxifene hydrochloride as a therapeutic molecule, that is known to decrease osteoclast activity and therefore enhanced bone formation. From histological and scanning electron microscopy evaluations, the use of Bonelike® associated with Floseal® and raloxifene hydrochloride showed that new bone was rapidly apposed on implanted granules and also that the presence of the matrix and therapeutic molecule does not alter the proven highly osteoconductivity properties of Bonelike®. Therefore, this association may be one step-forward for the clinical applications of Bonelike® scaffolds since it is much more easy-to-handle when compared to granular materials.     

Homogenization of hardness distribution and distortion in high pressure gas quenching

Quenching with gases rather than oil or other liquid media has the advantages of reducing the risks concerning health and environment, while simultaneously homogenizing the quenching results and minimizing distortion due to a wide range of possible process parameter variations and the pure convective heat transfer. In this contribution, a coupled solution for increasing homogenization of quenching results within high pressure gas quenching will be presented. In the first stage, an experimental test facility was set up for flow investigations and in the second stage a numerical simulation model was generated. The numerical and experimental results of the flow through the chamber were compared for several boundary conditions. Finally, after complete verification of the simulation, the model may be used to assist in parameter variation for optimization of homogeneous high pressure gas quenching.