湿法炼锌赤铁矿法沉铁行为

湿法炼锌工艺过程中赤铁矿法沉铁技术是一种沉淀渣量小、分离效率高的锌、铁分离方法。针对湿法炼锌工艺过程中赤铁矿法沉铁技术开展了硫酸亚铁的氧化水解沉淀行为的研究,考察了时间、酸度、温度、硫酸盐浓度、晶种返回量等主要因素对赤铁矿法沉铁效果的影响、沉淀产物的析出特性和物相表征。     

学术共同体理论研究综述

学术的标准、学术上的分歧等等,所有这些学术上的问题只有依靠学术共同体才有可能得到解决。同时发展学术共同体对促进科学发展,加强学术交流与合作,培养高级专业人才,提高学术审查与评价的权威等方面具有重要意义。加之,学术共同体是一个使用频率很高,却存在一定争议的概念。故此,加强学术共同体的研究是十分有必要的。     

SCFNN-Based Decision Feedback Equalization Robust to Frequency Offset and Phase Noise

This paper proposes a self-constructing fuzzy neural network-based decision feedback equalizer (SCFNN DFE). An online learning algorithm containing the structure and parameter learning phases is employed in training the SCFNN DFE. Specifically, the feedforward input vector classification and a gradient-descent method are both used in this online learning algorithm. We show by simulations that the proposed SCFNN DFE offers improvement compared to the traditional DFE methods in the presence of frequency offset and phase noise.     

A flip-chip packaged 80-nm In0.7Ga0.3As MHEMT for millimeter-wave low-noise applications

In this paper, we present a flip-chip 80-nm In_0.7Ga_0.3As MHEMT device on an alumina (Al₂O₃) substrate with very little decay on device RF performance up to 60 GHz. After package, the device exhibited high I_DS = 435 mA/mm at V_DS = 1.5 V, high g_m = 930 mS/mm at V_DS = 1.3 V, the measured gain was 7.5 dB and the minimum noise figure (NF_min) was 2.5 dB at 60 GHz. As compared to the bare chip, the packaged device exhibited very small degradation in performance. The result shows that with proper design of the matching circuits and packaging materials, the flip-chip technology can be used for discrete low noise FET package up to millimeter-wave range.     

A Design of Linearity Built-in Self-Test for Current-Steering DAC

In this paper, a current-mode Built-In Self-Test (BIST) scheme is proposed for on-chip estimating static non-linearity errors in current-steering digital-to-analog converters (DACs). The proposed DAC BIST scheme is designed to verify a 10-bit segmented current-steering DAC, consist of a 5-bit coarse DAC and a 5-bit fine one. This proposed BIST scheme includes a current-mode sample-and-difference circuit to increase the sampling current accuracy and control a current-controlled oscillator (ICO). In addition, only 36 measurements are required by using the selected-code method rather than 1024 measurements for the conventionally-utilized all-code method. Compared to the conventionally-utilized all-code method, about 85-% reduction of test time can be achieved.     

油菜秸秆在碳酸乙烯酯中液化工艺的研究

利用碳酸乙烯酯(EC)为液化剂在常压条件下对油菜秸秆进行了液化试验。选择甲烷磺酸(MSA)作为最优催化剂,探讨了反应温度、反应时间、物料量、催化剂量4个因素对液化得率及黏度的影响。采用二次回归正交旋转组合设计对其液化工艺进行了优化。结果表明:最佳液化条件为反应温度165℃,反应时间70 min,物料量15.5%,催化剂量3.2%。在此条件下,液化得率可达到92.89%。     

多特征融合的行为识别模型

目的 视频行为识别和理解是智能监控、人机交互和虚拟现实等诸多应用中的一项基础技术,由于视频时空结构的复杂性,以及视频内容的多样性,当前行为识别仍面临如何高效提取视频的时域表示、如何高效提取视频特征并在时间轴上建模的难点问题。针对这些难点,提出了一种多特征融合的行为识别模型。方法 首先,提取视频中高频信息和低频信息,采用本文提出的两帧融合算法和三帧融合算法压缩原始数据,保留原始视频绝大多数信息,增强原始数据集,更好地表达原始行为信息。其次,设计双路特征提取网络,一路将融合数据正向输入网络提取细节特征,另一路将融合数据逆向输入网络提取整体特征,接着将两路特征加权融合,每一路特征提取网络均使用通用视频描述符——3D ConvNets （3D convolutional neural networks）结构。然后,采用BiConvLSTM （bidirectional convolutional long short-term memory network）网络对融合特征进一步提取局部信息并在时间轴上建模,解决视频序列中某些行为间隔相对较长的问题。最后,利用Softmax最大化似然函数分类行为动作。结果 为了验证本文算法的有效性,在公开的行为识别数据集UCF101和HMDB51上,采用5折交叉验证的方式进行整体测试与分析,然后针对每类行为动作进行比较统计。结果表明,本文算法在两个验证集上的平均准确率分别为96.47%和80.03%。结论 通过与目前主流行为识别模型比较,本文提出的多特征模型获得了最高的识别精度,具有通用、紧凑、简单和高效的特点。     

Characteristic of copper wire and transient analysis on wirebonding process

In the present paper, two major analyses are achieved. In the first, experimental procedures were accomplished to measure tensile mechanical properties of copper (Cu) wire (? = 1 mil) before/after electric flame-off (EFO). Characteristics of free air ball (FAB), heat affected zone (HAZ) and thermal stable zone (TSZ) in as-drawn wire have been carefully investigated by microhardness, self-design pull test fixture, nanoindentation and atomic force microscopy (AFM). A 2nd EFO real-time technique has been conducted to reduce the strength of Cu wire and increase the bonding region. Secondary, with the obtained experimental material data, a comprehensive finite element wirebonding model based on explicit time integration software ANSYS/LS-DYNA is developed to predict the overall strain/stress distributions on the aluminum (Al) bond pad. Finite element analysis (FEA) results demonstrate that plastic deformation on Al bond pad around smashed FAB can be reduced by increasing the surface roughness on FAB. A series of comprehensive parametric studies were conducted in this research.     

Improving an LDMOST by variation of lateral doping on epitaxial-layer drift region

Variation of lateral doping (VLD) is applied to the epitaxial-layer drift region in LDMOSTs, resulting in non-uniform doping concentration. In addition to a decrease in on-resistance, it improves electric field distribution on the surface of drift region, and thus increases the breakdown voltage. We studied the condition of using the VLD technique, and compared it with the conventional technique of optimizing the epitaxial-layer uniform concentration. Results from TSPUREM 4 and MEDICI simulations indicate that when the epitaxial-layer doping concentration is 1.5 × 10¹⁵ cm⁻³, applying VLD to the drift region increases the breakdown voltage by as much as 34% (187-251 V), while the specific on-resistance is lowered by 55% (49-22 mΩ cm²), which are remarkable improvements. On the other hand, if the doping concentration of epitaxial-layer is already high, the improvement of applying VLD is limited.     

Fabrication of AD/DA microfluidic converter using deep reactive ion etching of silicon and low temperature wafer bonding

The purpose of this work is to describe an original process that has been designed for the fabrication of a microfluidic converter. The fabrication is based on deep reactive ion etching of silicon and low temperature full wafer adhesive bonding. The technology development includes an improvement of the bonding process in order to produce an adaptive strength of SU-8 bond which not only ensures absence of debonding failures during the silicon deep etching procedure and the subsequent dicing procedure, but also avoids the potential SU-8 overflow leakage into channels due to the bonding step. Besides, the originality of the work is not only in the process but also in the design of the device. Common actuation method for microfluidic system is either based on closed-channel continuous-flow microfluidic (CMF) or droplet-based microfluidic (DMF). Both of them have advantages and disadvantages, and their integration on a single system is in dire need. In this paper, we briefly discuss the concept of microfluidic converter, integrating CMF with DMF, which can: (i) continuously preload reagents, (ii) independently manipulate several droplets, (iii) recombine and export samples into closed-channel continuous flow, making it ideal for interfacing to liquid-handling instruments and micro-analytical instruments.