材料期刊网，材料期刊，材料类期刊，材料期刊投稿

本文论证了引入模式识别技术的可能性和必要性.简单介绍了模式识别技术,重点叙述了非线性映照,主要量分析,Fisher判别,聚类分析等方法的数学原理和应用特点.根据实际应用的经验,讨论了数据预处理,训练集选择,特征参数和关键参数,以及方法效果验证方法等一系列基础问题,并介绍了模式识别技术在环境腐蚀科学中的部分应用实例.

关键词: 环境腐蚀科学 , pattern recognition , application , review

污染离子对Q235钢潮湿时间的多因素关联性分析

唐子龙李超曲文超宋鑫

中国腐蚀与防护学报

研究了Q235钢在1 mmol/L NaCl，10 mmol/L NaSO₄和NaCO₃薄液膜下潮湿时间随温度、相对湿度和浸润时间的变化规律。采用数据挖掘中标准化技术对自变量和因变量进行预处理，提高了分析精度。通过线性模型中各自变量的权重系数研究环境控制因素对潮湿时间的影响规律和贡献份数，分析污染离子和腐蚀产物对潮湿时间的作用方式。结果表明：数据标准化技术可有效提高分析精度。 Cl^-、SO₄^2-和CO$_₃^2-三种污染离子对于材料表面潮湿时间有不同的影响模式。温度和湿度对潮湿时间的影响较大，浸润时间对潮湿时间的影响较小。浸润时间和湿度跟潮湿时间呈正相关，而温度跟潮湿时间呈负相关。在24 ℃和相对湿度为70%~80%左右(转化点)温度和相对湿度指数有显著变化。温度和相对湿度对潮湿时间的影响有明显负协同作用，强弱和模式依赖于污染离子种类和理化性质，与腐蚀产物也有密切关系。

关键词: 大气腐蚀 , thin liquid layer , time of wetness , data mining , pattern recognition

AN EXPERT SYSTEM FOR COMPUTER AIDED ALLOY STEEL DESIGN—ASPRES

CHEN Yu , CHEN Nianyi Shanghai institute of Metallurgy , Academia Sinica , Shanghai , China PENG Ping , KONG Bingyu Iron and Steel Research Institute , Shanghai Baoshan Iron and Steel Complex , Shanghai , China professor , Shanghai Institute of Metallurgy , Academia Sinica , Shanghai 200050 , China

金属学报(英文版)

An alloy steel pattern recognition expert system,ASPRES,has been established for the purpose of computer aided optimal design of alloy steel in compositions and process pa- rameters,Pattern recognition techniques are used to abstract inner relationship between mechanical properties and process variables.The ASPRES uses 2-dimensional graph as visual knowledge to represent domain expertise of specific object.Forward and back- ward chaining can be utilized by researcher in predicting sample performances or giving helpful suggestions about the chemical compositions and process parameters according to desired properties.

关键词: pattern recognition , null , null

基于提升小波的矩不变量车牌字符识别方法

高强 , 刘斌

量子电子学报 doi:10.3969/j.issn.1007-5461.2016.06.004

车牌识别系统的关键在于字符的识别,字符识别的核心是提取字符特征.小波变换可以获取字符的细节结构特征,不变矩能很好地对其进行描述,将两者结合起来提取字符的特征.利用张量积小波分解高频子图具有方向性的特点,提取字符的笔画特征,得到反映字符结构和统计特征的联和特征向量.字符图像的分解采用第二代提升小波算法,进一步降低了计算复杂性.实验结果表明,此方法提取得到的联合特征向量能达到98％的字符识别率,可以满足实际应用的要求.

关键词: 信息处理 , 模式识别 , 提升小波变换 , 不变矩 , 统计特征 , 结构特征 , 笔画特征

FACTORS AFFECTING SILICA CONTENT OF ALUMINIUM HYDROXIDE PRODUCT IN CARBONIZATION PROCESS

LIU Miaoxiu CHENG Zhaonian CHEN Nianyi Shanghai Institute of Metallurgy , Academia Sinica , Shanghai , ChinaKE Jiajun Institute of Chemical Metallurgy , Academia Sinica , Beijing , China CHEN Nianyi , Professor , Shanghai Institute of Metallurgy , Academia Sinica , Shanghai 200050 , China

金属学报(英文版)

The regularity of the carbonization process in alumina production has been summarized by the pattern recognition and regression.A mathematical model has been proposed to describe the factors affecting the silica content of aluminium hydroxide product.The high concentration of aluminate ions in solution will produce aluminium hydroxide with low content of SiO_2 under the condition of Al_2O_3/SiO_2 ratio kept unchanged.This was confirmed by certain scale experiments.

关键词: carbonization , null , null , null

PATTERN RECOGNITION APLLIED TO STABILITY OF FILLED Ti_2Ni PHASES

ZHOU Diangen CHEN Nianyi Shanghai Institute of Metallurgy , Academia Sinica , Shanghai , China professor , Shanghai Institute of Metallurgy , Academia Sinica , Shanghai , 200050 , China

金属学报(英文版)

Pattern recognition method is used for the investigation of stability,region of filled Ti_2Ni phases in multi-dimensional bond-parameter space.The filling of C,N and O atoms into T_6 octahedra consisting of atoms of earhy-transition elements makes the expansion of the stability region of Ti_2Ni phase,and the relative stability of AI_2Cu and MoSi_2 type com- pounds decreases after the introduction of non-metallic elements such as C,N and O.

关键词: filledTi_2Ni phase , null

FACTORS AFFECTING SILICA CONTENT OF ALUMINIUM HYDROXIDE PRODUCT IN CARBONIZATION PROCESS

LIU Miaoxiu CHENG Zhaonian CHEN Nianyi Shanghai Institute of Metallurgy , Academia Sinica , Shanghai , ChinaKE Jiajun Institute of Chemical Metallurgy , Academia Sinica , Beijing , China

金属学报(英文版)

The regularity of the carbonization process in alumina production has been summarized by the pattern recognition and regression.A mathematical model has been proposed to describe the factors affecting the silica content of aluminium hydroxide product.The high concentration of aluminate ions in solution will produce aluminium hydroxide with low content of SiO_2 under the condition of Al_2O_3/SiO_2 ratio kept unchanged.This was confirmed by certain scale experiments.

关键词: carbonization , null , null , null

THE FORMABILITY OF COMPLEX FLOURIDE COMPOUNDS

LI Zheng , LIU Honglin , CHENG Zhaonian , CHEN Nianyi , Shanghai Institute of Metallurgy , Academia Sinica , Shanghai , ChinaZHOU Jiaju , Institute of Chemical Metallurgy , Academia Sinica , Beijing , China CHEN Nianyi , Professor , Shanghai Institute of Metallurgy , Academia Sinica , Shanghai 200050 , China

金属学报(英文版)

The formability of complex flourides has been studied by chemical bond parameter-pattern recognition method. In a multi-dimensional space spanned by ionic radius, ionic polarizability, electronegativity and ionic charge of the metal elements, the binary complex flouride-forming systems and binary flouride systems without complex flouride formation distribute in different regions with a clear-cut boundary. The mathematical models obtained are useful for new complex flouride predicton.

关键词: complex flouride , null , null

智能识别方法在板形识别中的应用及发展趋势

张雪伟 , 王焱

钢铁研究学报

对国内外关于板形模式识别技术的研究现状和发展趋势进行了综述.通过比较,提出了传统的基于最小二乘法的多项式拟合法存在的不足,并对模糊分类、神经网络、遗传算法、混沌优化等智能识别方法在板形模式识别中所具有的优势进行了归纳和总结.最后,对智能方法在板形识别问题中的应用以及板形识别技术的发展趋势进行了展望,为板形检测环节得到理想板形信号提供理论研究方法,并将逐步应用于钢铁板形控制的工业过程中.

关键词: 板形 , 模式识别 , 智能方法 , 优化