基于层流流量计的航发燃油流量动态测量技术研究

为解决航空发动机过渡态燃油测量问题,研究了基于层流流量计(LFM)的航空燃油流量动态测量技术.设计了一款量程为0～3 000 mL/min的液体LFM,在小型涡轮喷气式发动机试车台上对其进行了试验,并与科里奥利流量计(CMF)进行了比对,试验结果显示LFM具有优秀的动态特性.当发动机处于稳态时,LFM测量结果与CMF吻...     

基于S变换的电能质量扰动支持向量机分类识别

采用s变换和支持向量机进行电能质量扰动的分类识别。作为连续小波变换和短时傅立叶变换的发展，S变换引入了宽度与频率成反向变化的高斯窗，具有与频率相关的分辨率。由于S变换具有良好的时频特性，因而非常适合于进行电能质量扰动信号特征提取。首先通过S变换进行扰动信号特征提取，然后构造支持向量机分类树进行扰动分类。算例表明该方案具有分类准确率高，对噪声不敏感，训练样本少等优点，是电能质量扰动识别的有效方法。     

锅炉炉膛及烟气系统防爆设计压力取值标准的分析

锅炉炉膛及烟气系统防爆设计压力尤其是最大设计负压的取值,对锅炉及烟风系统的设计、燃煤机组运行的安全性和经济性都有重要影响,但目前防爆设计规范中有关的规定存在着明显不足.通过对锅炉主燃料跳闸( main fuel trip,MFT)动态特性和引风机匹配特性分析,论证了按NFPA规范对炉膛瞬态设计负压绝对值通常不必超过8.7 kPa的合理性,并推导得到了对烟气系统防爆设计负压的计算方法和安全性评估法则,为规范炉膛及烟道设计压力的标准提供参考.     

三维流体力学预测风机不同布置形式对冷库气流的影响

我国冷库的建设数量近年来增长迅速,可多数冷库中的气流分布并不合理,而风机的布置形式是影响冷库中气流分布的一个重要因素。运用SIMPLE算法,结合波兴涅克（Boussineq）假设,对目标冷库中不同的冷风机安装形式进行三维数值模拟。模拟结果显示目标冷库中采用两个三出风口的风机吹风比采用三个两出风口的风机吹风所引起的气流场均匀性更好;同时针对后者造成的射流偏移,对风机位置进行改变后得出了冷库中的流场较改变前均匀性提高。数值模拟的结果对中型冷库和大型冷库中风机的安装形式和安装位置提供了参考依据。     

Microwave puffing: Determination of optimal conditions using a coupled multiphase porous media - Large deformation model

Increased interest in microwave puffing is due to its ability to obtain low-fat and ready-to-eat healthy products. Determination of optimal conditions for this complex process has been difficult and although several patents exist on the concept, we are yet to see any large scale commercial use. A fundamental physics based modeling approach integrated with relevant experimentation, developed in this work, is an ideal framework to understand and optimize microwave puffing. The results showed that puffing may not be successful unless carried out using an intensive heating source such as microwaves. Addition of infrared and hot air leads to better quality product whereas using forced air convection is not desirable. There is an optimum initial moisture content depending on the puffing conditions. The study provides critical guidelines to food product/process developers for successful development, control and automation of microwave puffing, thereby leading to value-added nutritious products.     

羊胴体产量分级模型初探

选择发育正常、健康无病的40只宁夏滩羊进行产量级6个指标的测量,采用主成分分析方法对其进行多元分析,结果表明:保留一个主成分方差累计贡献率达80.5%,且主成分内部显著相关,肋肉厚和腹壁厚的相关系数为0.97201,最终确定羊胴体产量分级指标为肋肉厚。对其与产肉率进行模型拟合,得到羊胴体产量分级的最佳模型,模型决定系数达到0.9614。     

四川地方晾晒烟品种叶面形状分析

为分析四川地方晾晒烟烟叶叶面形状特点,在烟叶成熟期测量和计算了50个品种的中部烟叶的叶尖到叶基的长度、叶片最宽处的宽度、叶基到叶长线与宽线交点处的长度、叶尖到叶长线与宽线交点处的长度、叶尖角度、叶面积等9个叶面形状描述指标,通过主成分分析和聚类分析建立了多指标叶面形状描述法。该方法可将50个供试品种的叶面形状划分为宽卵形、心脏形、长卵形、卵形、椭圆形、长椭圆形、宽披针形、披针形8个类型,以椭圆形、卵形及它们的变型为主,占全体样本的80%,可有效地用于烟叶叶面形状的描述及分类。     

Effects of buffer and temperature on formation of furan, acetic acid and formic acid from carbohydrate model systems

Effects of temperature, pH and phosphate buffer on volatiles formed by heating model solutions of fructose, glucose, sucrose, lactose or starch were investigated by selected ion flow tube-mass spectrometry (SIFT-MS). Among the carbohydrates studied, the monosaccharide reducing sugars fructose and glucose were more reactive than macro molecular starch, lactose and sucrose, which are not reducing sugars. Furan was formed from fructose heated at 120, 100 and 80 °C at pH 5-7 in phosphate buffer and at 100 °C in unbuffered solution. Glucose did not yield furan in unbuffered solution at 100 °C but glucose did yield furan in phosphate buffer solution at a similar pH. Furan formation increased as temperature and pH increased. As temperature, pH value, and buffer components changed, furan formation may occur by different reactive pathways. Both of the buffer components, NaH₂PO₄ and citric acid, enhanced the formation of thermal degradation products from fructose. The enhancing effect of NaH₂PO₄ was stronger than citric acid. Formic and acetic acids were concomitantly formed with furan. More formic acid was produced from fructose than glucose, and more acetic acid was produced from glucose than fructose.     

亚洲帘线钢线材生产现状及发展趋势

介绍亚洲帘线钢线材生产现状:日本新日铁开发了DLP控冷工艺保证帘线钢产品的高质量;日本神户制钢开发的超洁净钢冶炼工艺,很好地控制了钢中P,S,O的含量,以及夹杂物的含量、性质,使得神户的帘线钢线材成为世界最优;韩国浦项通过优化斯太尔摩冷却工艺来生产高品质帘线钢线材;宝钢帘线钢经历了转炉模铸、电炉小方坯、电炉大方坯连铸工...     

基于SVM方法的猪肉新鲜度分类问题研究

文中采取减压贮藏方式对新鲜猪肉进行了贮藏实验,测定了不同贮藏时间样品的挥发性盐基氮含量(TVB-N)、细菌总数、pH值及感官评价数据,并运用支持向量机(support vector machine,SVM)对这些样本数据进行训练,选取不同的核函数,得到SVM神经网络模型,随后利用此模型对测试数据进行猪肉新鲜度分类预测。实验表明,根据样本特性进行数据预处理,且选取合适的核函数后,SVM神经网络能得到极高的猪肉新鲜度正确分类率。