复杂系统建模——高维特征空间变量法

本文运用模式识别技术提出了一种分析和选择复杂系统变量的方法.运用该方法使得系 统的建模和控制简化.本文以复杂碳化过程作为实例,用模式识别技术对其内在规律进行分 析,找出了温度分布与结晶质量的定量关系.在此基础上,建立了碳化过程的动态模型.实践 表明,该模型与实际过程基本吻合,并已取得明显经济效益.     

基于宽带毫米波工作体制的雷达目标识别方法探讨

对基于宽带毫米波工作体制的雷达目标特征获取和识别方法进行了综合分析和探讨.表明:在毫米波照射下,雷达目标可模型化为由多个散射中心组成的扩展目标;只要发射信号有足够的带宽,就可以从回波信号中获得各散射中心的空间分布、类型等特征;仿真结果说明,这种方法对高频区复杂目标的识别是有效的.     

模式识别催化剂生产调优

本文介绍如何将模式识别应用于生产调优,着重讨论了多指标(优化目标)因素的分级,提出了用模糊综合评价度量和依有序聚类最小损失函数准则划类,并相继进行变量与样本筛选、信息压缩、特征提取和模拟仿真获得优区操作条件,实施效果显著。     

基于HMM方法的动态手势轨迹训练性能研究

基于HMM(Hidden Markov Model,隐形马尔可夫模型)对动态手势轨迹的训练是手势识别的关键技术之一。本文对HMM的模型训练采用Baum.Welch算法,并分别从迭代次数,样本个数选取,以及模型初值选取等方面对动态手势轨迹的训练性能进行了研究。实验结果表明HMM方法对具有时空特性的动态手势轨迹识别是非常有效的。     

运用模式分类的雷达抗转发式距离欺骗干扰方法

针对雷达转发式欺骗干扰中的距离欺骗，提出了一种基于模式分类来识别目标与干扰的方法。该方法基于干扰和目标能量和起伏特性差异确定提取特征因子，再利用RBF神经网络方法进行分类。仿真结果表明，该方法具有较好的抗距离欺骗干扰性能。     

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。     

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…     

由三面投影图生成三维立体图的方法研究

针对创新设计中零部件的三维虚拟造型问题，提出了一种由二维投影图生成三维立体图的计算机化辨识方法。通过辨识简单三维立体的形状和空间位置，克服了二维图形中面的积聚性和零件之间不规则相贯和截交曲线的识别问题，进而实现了复杂组合形体的计算机化设计与造型。最后，以箱体类零件为例，在AutoCAD环境下利用VC 和ObiectARX开发工具编程验证了该方法的有效性。     

Photocatalytic degradation of 2,2‐bis(4‐hydroxy‐3‐methylphenyl) propane (BPP) based on molecular recognition interaction

BACKGROUND: Endocrine disruptors in the aquatic environment and their potential adverse effects are currently issues of concern. One of these endocrine disruptors is 2,2‐bis(4‐hydroxy‐3‐methylphenyl)propane (BPP). In this work the molecular recognition interaction of BPP with β‐cyclodextrin (β‐CD) was studied using IR spectroscopy and steady state fluorescence spectroscopy, and the photocatalytic degradation behaviour of BPP based on molecular recognition interaction was investigated in a TiO₂/UV–visible (λ_max = 365 nm) system. This might provide a new method for the treatment of some organic pollutants in wastewater. RESULTS: β‐CD reacts with BPP to form a 1:1 inclusion complex, the formation constant of which is 4.94 × 10³ L mol^?1. The photodegradation rate constant of BPP after molecular recognition by β‐CD showed a 1.42‐fold increase in the TiO₂/UV–visible (λ_max = 365 nm) system. The photodegradation of BPP depended on the concentration of β‐CD, the pH value, the gaseous medium and the initial concentration of BPP. The photodegradation efficiency of BPP with molecular recognition was higher than that without molecular recognition. After 100 min of irradiation the mineralisation efficiency of BPP after molecular recognition by β‐CD reached 94.8%, whereas the mineralisation efficiency of BPP before molecular recognition by β‐CD was only 40.6%. CONCLUSION: The photocatalytic degradation of BPP after molecular recognition by β‐CD can be enhanced in the TiO₂/UV‐visible (λ_max = 365 nm) system. This enhancement is dependent on the enhancement of the adsorption of BPP, the moderate inclusion depth of BPP in the β‐CD cavity and the increase in the frontier electron density of BPP after molecular recognition. Copyright © 2008 Society of Chemical Industry