零无功滤波器的设计与工程应用

针对目前部分末端用户负荷功率因数较高,而谐波严重的情况,研究设计了一种新的电路,使滤波器运行时输出无功较小,对于特定的谐波其阻抗接近于零,因此达到滤波的目的,并将该技术首次应用于温州文成钢厂的实际工程中,效果良好。     

电网谐波治理相关问题探讨

对谐波相关国家标准的可操作性、供电企业的责任、谐波用户管理及谐波治理方案的选择等问题进行了探讨,并提出了有关建议,对电网的谐波治理工作有一定的参考意义。     

基于Leap Motion的动态手势识别方法

伴随虚拟现实（Virtual Reality,VR）技术的发展,以及人们对人机交互性能和体验感的要求提高,手势识别作为影响虚拟现实中交互操作的重要技术之一,其精确度急需提升^[1].针对当前手势识别方法在一些动作类似的手势识别中表现欠佳的问题,提出了一种多特征动态手势识别方法.该方法首先使用体感控制器Leap Motion追踪动态手势获取数据,然后在特征提取过程中增加对位移向量角度和拐点判定计数的提取,接着进行动态手势隐马尔科夫模型（Hidden Markov Model,HMM）的训练,最后根据待测手势与模型的匹配率进行识别.从实验结果中得出,该多特征识别方法能够提升相似手势的识别率.     

高速动车组网侧电流谐波特性的研究

研究了高速动车组注入牵引网电流的谐波特性。分析了两电平及三电平4象限变流器的主电路拓扑及控制策略。利用双边傅立叶级数法,给出了两种4象限变流器网侧电流的谐波公式,总结了谐波分布规律。编制了CRH2和CRH5型动车组4象限变流器的仿真软件,研究了不同牵引和制动功率下,网侧电流的THD及各次谐波电流大小,验证了理论分析的正确性。     

多分辨HHT在STATCOM并补线路故障位置识别中的应用

引入多分辨分析方法对Hilbert-Huang变换(HHT)进行改进:根据信号的特征尺度对信号筛分后进行经验模态分解(EMD)处理,然后对固有模态函数(IMF)分量进行Hilbert变换,求取各分量的瞬时频率和幅值。通过对信号按不同时间尺度进行滤波,有效消除模态混叠;同时对信号进行分段样条拟合,减少计算量。用多分辨HHT方法对含静止同步补偿器(STATCOM)的并补线路保护在稳态和故障暂态时的测量信号以及在不同故障位置时的谐波进行仿真分析,结果表明:STATCOM会使故障时保护安装处的电压谐波含量明显增大;谐波分量的特征与故障回路中是否包含STATCOM有关,故障回路包含STATCOM时特征谐波的幅值很大,否则幅值很小。将该谐波特征用于STATCOM并补线路的故障位置识别,可以自适应地判断故障点的位置,与距离保护相结合,可以有效地解决距离保护用于并补线路时的超越误动和欠范围拒动问题。     

A mathematical model for the distribution balance of extraction processes from waste batteries

The experiments of metal Cd extraction were carried out three times at oil phase/aqueous phase (O/A)=1 and pH 2.2 by using 1.3 mol/L di(2-ethylhexyl)phosphoric acid (DEHPA) as the agent,with an extraction rate of 99.5%.Experiments on metal Co extraction were done three times at O/A=0.5 and pH 4.7 by using 0.4 mol/L Cyanex 272 as the agent,with an extraction rate of 99.2%.By using equilibrium data and massive calculations to analyze and predict the interaction among the extraction systems,this paper came out with the polynomials of the empirical model as well as their parameter values for the equilibrium concentration of the above two extraction loops.     

Use of mesophilic and thermophilic bacteria for the improvement of copper extraction from a low-grade ore

Bioleaching was examined for copper extraction from a low grade ore using mesophilic and moderate thermophilic bacteria. Five equal size columns were used for the leaching of the ore. Sulfuric acid solution with a flow rate of 3.12 L.m-2.h-1 and pH 1.5 passed through each column continuously for 90 d. In the first and the second column, bioleaching was performed without agglomeration of the ore and on the agglomerated ore, respectively. 28wt% of the copper was extracted in the first column after 40 d, while this figure was 38wt% in the second column. After 90 d, however, the overall extractions were almost the same for both of them. Bioleaching with mesophilic bacteria was performed in the third column without agglomeration of the ore and in the fourth column on the agglomerated ore. After 40 d, copper extractions in the third and the fourth columns were 62wt% and 70wt%, respectively. Copper extractions were 75wt% for both the columns after 90 d. For the last column, bioleaching was performed with moderate thermophilic bacteria and agglomerated ore. Copper extractions were 80wt% and 85wt% after 40 and 90 d, respectively. It was concluded that crushing and agglomeration of the ore using bacteria could enhance the copper extraction considerably.     

基于深度学习的采摘机器人目标识别定位算法

为了提升采摘机器人对果实的识别准确率以及定位定精度,提出一种基于深度学习Faster-RCNN框架的采摘机器人目标识别和定位算法.首先采用卷积神经网络VGG16模型提取输入图像的特性信息,并利用区域提议网络RPN生成含有目标的候选框,通过引入自适应候选框数的方法有效提升了算法性能,然后利用多任务损失函数对目标进行分类识...     

海水卤水提锂高效吸附剂的合成及应用研究

高效吸附剂的研究和开发是高镁锂比海水卤水中锂分离提取的关键,当前最具工业应用前景的吸附剂为锂锰氧化物离子筛。本研究探讨了各合成条件对锂锰氧化物离子筛吸附剂的结构及吸附性能的影响,合成出了适用于不同镁锂比及锂浓度海水卤水的高容量吸附剂,吸附容量最高可达41.02 mg/g,锰溶损率为3.06%;此外,还开展了锂吸附分离的动态模拟试验,为所合成吸附剂用于海水卤水提锂的应用研究奠定基础。