黄埔发电厂4号机组协调控制系统的改造

黄埔发电厂4号机组控制系统运行情况一直不理想，影响到机组的安全运行，简要介绍了热工人员通过对原系统存在的几种常见缺陷进行的分析和研究，所提出的一套合理的改造方案付诸于实践，取得了较好的效果。     

外源多糖水解酶提高红碎茶品质的生化机制研究

本文在肯定红碎茶初制中外源多糖水解酶（PSH）增质效果的基础上,利用电镜及高效液相色谱等分析手段对外源PSH的增质机理进行了探讨,研究结果表明：外源PSH加工红碎茶的水解效果明显,可使还原糖和水浸出物大幅度提高,且发酵过程中存明显的水解优势期,外源PSH水解叶组织细胞壁中的不溶出性多糖产生生化破损作用,使发酵茶胚酸化,提高多酚氧化酶活性,促进茶黄素形成及叶绿素降解,同时提高成茶有效成分的浸出率,从而较全面地改善红碎茶的内外品质,PSH水解作用所致的细胞生化破损效应是外源PSH综合品质效应的核心因素。     

The correlation between skin-care effects and phytochemical contents in Lamiaceae plants

In this study, well-known skin-care functional parameters were used, including UV absorption, DPPH-scavenging, NO-production-inhibition, tyrosinase-inhibition activity and anti-Staphylococcus aureus activity to measure the effects of the 70% acetone-extracts form 28 species of Lamiaceae plants. Further, the phytochemical contents were explored by total phenol (TP), total flavone (TF) and total coumarin (TC). The correlation between the skin-care effects and the phytochemical contents was analysed by non-parametric correlation analysis. Amongst the 70% acetone-extracts, Origanum majorana displayed the strongest DPPH-scavenging and tyrosinase-inhibitory effects and the richest phenol content. Based on the statistics results, the phytochemical contents were related with those parameters, such as: DPPH-scavenging effects vs. TP (R = 0.542), TF vs. TP (R = 0.613), and NO-inhibitory vs. anti-bacterial activities (R = 0.767). Moreover, each genus of Lamiaceae had different properties of skin-care effects. Form our research works, the Lamiaceae is good resources to develop skin-care cosmetics.     

TT 系统在室外配电系统中的应用研究

本文分析了TT系统的主要优势,阐述了TT系统中安装剩余电流保护器的必要性,并且通过分析计算给出了在安装剩余电流保护器的情况下,TT系统中设备保护接地的接地电阻应取值的范围以及接地电阻的具体设计方案。     

Antioxidant stability of small fruits in postharvest storage at room and refrigerator temperatures

Strawberries (Fragaria ananassa), raspberries (Rubus idaeus) and red currants (Ribes rubrum), as well as two drupes, cherries (Prunus avium), and sour cherries (Prunus cerasus), were subjected to two storage temperatures (4 °C and 25 °C) and phytochemicals concentrations (total phenols, flavonoids and anthocyanins) as well as antioxidant capacity (DPPH, ABTS and FRAP assays) were monitored until the fruit visually spoiled. Red currants and strawberries exhibited the highest initial total phenol (TP) contents (322.40 ± 5.56 and 335.47 ± 6.12 mg GAE/100 g FW, respectively) and maintained the highest TP contents throughout storage at both temperatures. Storage of at 25 °C as opposed to 4 °C, facilitated faster spoilage of analyzed fruits. In addition, most fruits stored at 4 °C, exhibited slightly higher antioxidant activity values at the end of storage according to all three antioxidant activity assays as opposed to fruits stored at 25 °C. The dynamic evolution of antioxidant capacity at both temperatures reflected the transient changes in phytochemical composition of small fruits in storage.     

响应传导函数在船舶机械噪声预报中的应用及实验验证

据线性时不变系统自身具备的响应传导特性,将力-响应传递函数模型转化为响应-响应传导函数模型用于船舶机械噪声预报。理论推导传导函数法用于响应预报的约束条件,分析其用于船舶机械噪声实时预报的可行性。给出传导函数的求解方法 -工况联合法,并对其应用条件进行分析论证。双层圆柱壳体模型试验结果表明,传导函数法用于近、远场机械噪声预报时均具有较高精确度。     

基于微博转发的社交网络模型

探究微博转发网络的构建机制有助于深刻理解信息在微博平台上的传播过程,得出针对微博营销、舆论管控的有效策略和建议。针对这一问题,提出了一种有向加权网络模型。根据微博在转发过程中被多次转发的现象,在模型建立过程中,在节点间连边时引入三角连接机制,并且用连边的方向选择表征了活跃用户和知名用户的不同行为特征,同时考虑了权值的动态演化过程。理论分析和仿真实验表明模型的强度分布、度分布、强度-度的相关性均服从幂律分布,幂指数为1~3,而且具有高聚类、短路径的特点,平均聚类系数可达0.42,平均路径长不超过6,同时采集了微博转发的实际数据验证了模型的正确性。     

基于欠定盲分离的同步/异步跳频网台分选方法

研究无线通信网台分选优化问题,现有的跳频网台分选大多要求阵元数必须大于信号数,由于在实际应用中常常很难满足,为了用有限的阵元数分选尽可能多的跳频信号,提出了一种基于欠定盲分离的跳频信号分选方法.根据跳频信号具有良好的时频稀疏性,依据稀疏分量分析的“两步法”思路,首先用单源点聚类方法估计混合矩阵;然后利用子空间投影方法分离各跳频信号,从而实现了跳频信号的欠定盲分离.仿真结果表明了改进方法的有效性,为通信网分选优化提供了参考.     

基于语音信号稀疏性的FDICA初始化和后处理方法

目前解决语音信号盲源分离(Blind source separation,BSS)的两大类方法分别为频域独立成分分析(Frequency domain independent component analysis,FDICA)和基于稀疏性的时频掩蔽(Time frequency masking,TF masking).为此将两类方法优点相结合,利用TF masking方法的结果,对FDICA做初始化,在加快FDICA收敛速度的同时也避免了次序不确定性问题.此外还提出了一种新的基于语音稀疏性FDICA的BSS后处理方法:基于局部最小比例控制(Local minimum ratio controlled,LMRC)谱减法,比常规的TF masking、维纳滤波等后处理方法,能够更有效地控制音乐噪声,提高分离性能.合成数据和实际采集数据的实验结果验证了所提方法的有效性.     

基于字同现频率的关键词自动抽取

为提高关键词自动抽取的准确率,提出了基于字同现频率的关键词自动抽取算法。根据词的位置和文本长度改进TF/IDF算法,由字同现频率计算词的信息量,运用特征加权计算词的权重,选取权重大的词作为关键词。给出了关键词自动抽取的过程,设计了关键词抽取的对比实验,验证该算法的有效性。实验结果表明该算法在准确率和召回率上具有优势。