新冠肺炎应急临床试验设计中的统计学考虑

新型冠状病毒肺炎的突然来袭，国内各地相继启动重大突发公共卫生事件一级响应。为了探索新冠肺炎的治疗，百余项临床研究陆续在国内开展。在突发公共卫生事件下，尽早找到有效的药物和治疗方案对于疾病的控制更加重要，这也对在这种应急情况下所开展临床试验的时效性提出要求，尤其是针对新冠肺炎的注册临床试验。本文从统计学角度，对应急临床试验中临床终点的选择和试验设计中的相关问题进行探讨。     

一种低信噪比环境下的语音端点检测算法

端点检测技术是语音信号处理的关键技术之一，为提高低信噪比环境下端点检测的准确率和稳健性，提出了一种非平稳噪声抑制和调制域谱减结合功率归一化倒谱距离的端点检测算法。该算法首先通过抑制非平稳噪声再采用调制域谱减消除残余噪声来提升信噪比，减少语音失真。然后再提取每帧信号的功率归一化倒谱系数，计算每帧信号与背景噪声的功率归一化倒谱距离。最后将该倒谱距离作为检测参数，采用双门限判决方法进行端点检测。实验结果表明，该端点检测算法对语音帧和噪声帧具有较好的区分性。此外，在低信噪比环境下，所提出的算法对于不同类型的噪声都具有较好的稳健性。     

基于改进型双门限语音端点检测算法的研究

语音端点检测是语音信号处理过程中的一个重要步骤,其检测准确性直接影响语音信号处理的速度和效果.传统的基于双门限法语言检测技术,在语音处于纯语音情况下判断语音端点较准确,但在语音处于噪声情况下,尤其是低信噪比的情况下,端点识别率很低,出错率很高.基于提高此方法识别率的目的,采用调整阈值个数,平滑滤波,引入语音结束最小长度的方法对其进行改进,通过了Matlab仿真实验,得出了较好的语音端点检测准确率.     

提高重轨钢终点碳含量的实践

通过对炼钢工艺的改进,结合攀枝花钢钒有限公司炼钢厂现有条件,实践确定将转炉冶炼重轨钢终点碳含量提高到0．1％～0．2％的范围可行,结果表明钢水终点磷、氧含量大幅降低,终点碳质量分数提高到0．145％,终点磷质量分数降低到0．006％,可以满足钢厂生产高品质重轨钢的需要。     

一种新的排序算法——端点排序算法

提出了一种新的排序算法：端点排序算法。其方法为：依次找出数据总数为N的数列最小和最大值,把二者放在本次所排数列的两端,再把剩余两端之间的数据总数为N-2的数列的最小值和最大值找出,放在此数列的两端,依此类推,直至数列中间,实现整个数组的排序。实验表明,该算法具有与冒泡排序更快的性能。在数据个数较多的情况下优于选择排序。     

Model analysis and experimental investigation of the friction torque during the CMP process

A model for calculating friction torque during the chemical mechanical polishing(CMP) process is presented,and the friction force and torque detection experiments during the CMP process are carried out to verify the model.The results show that the model can well describe the feature of friction torque during CMP processing. The research results provide a theoretical foundation for the CMP endpoint detection method based on the change of the torque of the polishing head rotational spindle.     

二苯胺磺酸钠指示剂终点颜色变化的探讨

以二苯胺磺酸钠作指示剂的重铬酸钾测定法是一种常用的容量分析法,主要用来测定样品中铁的含量,但在测定时,有时终点颜色得不到正常的紫色,而是出现墨黑色.文章对此现象进行了探讨并提出了一些解决方案.     

新型碱性精抛液对300mm铜膜全局平坦化的影响

The copper removal rate and uniformity of two types copper slurries were investigated, which was performed on the 300 mm chemical mechanical planarization （CMP） platform. The experiment results illustrate that the removal rate of the two slurries is nearly the same. Slurry A is mainly composed ofa FA/OI1 type chelating agent and the uniformity reaches to 88.32%. While the uniformity of slurry B is 96.68%, which is mainly composed of a FA/OV type chelating agent. This phenomenon demonstrates that under the same process conditions, the uniformity of different slurries is vastly different. The CMP performance was evaluated in terms of the dishing and erosion values. In this paper, the relationship between the uniformity and the planarization was deeply analyzed, which is mainly based on the endpoint detection mechanism. The experiment results reveal that the slurry with good uniformity has low dishing and erosion. The slurry with bad uniformity, by contract, increases Cu dishing significantly and causes copper loss in the recessed region. Therefore, the following conclusions are drawn： slurry B can improve the wafer leveling efficiently and minimize the resistance and current density along the line, which is helpful to improve the device yield and product reliability. This investigation provides a guide to improve the uniformity and achieve the global and local planarization. It is very significant to meet the requirements for 22 nm technology nodes and control the dishing and erosion efficiently.     

LEMD端点延拓方法及其在微震信号降噪中的应用

针对非平稳的微震信号,采用经验模态分解(Empirical Mode Decomposition,EMD)方法对其滤波、消噪。但该方法在应用中存在严重的端点效应,会影响分解结果。为了改善端点效应,针对微震信号低频特点,提出了基于LMS自适应滤波器对微震数据端点处进行波形延拓的方法。LMS自适应滤波器经过信号内部的数据训练后从端点处向左右延拓波形,延拓过程中生成符合内部信号趋势的极小值和极大值,波形的延拓长度由延拓算法自适应决定。在EMD对信号分解过程中得到前一层模态分量后均再次对剩余信号进行延拓,进而进行下一层分解,以此修正插值算法带来的误差。最后根据信号频率特点对得到的模态分量有选择的重构,达到降噪的目的。利用仿真数据和煤矿微震监测数据对LMS自适应延拓法进行了验证,结果表明该方法能够明显减小经验模式分解的端点效应,特别对低频的微震信号有很好效果。     

一种用于光纤周界安防系统的端点检测方法

为了提高光纤周界安防系统中扰动信号端点检测的 性能,提出一种新型端点检测方法。首先通过计算信号 的短时能量和过零率分布分别找出信号中短时能量比和过零率满足一定门限的信号点,通过 对这两信号点 进行一定的加权确定扰动信号的端点位置。与离散小波提取起始点信号相比,本文方法由于 综合了短时能量 和过零率法的优势,能够适应更多的扰动事件类型,通过加权极大提高了端点检测的精度 ,并且运算时 间大幅降低,提高了系统的实时性。实验结果表明,本文算法在10M Hz的采样速度和采集0. 3s数据情况下,93.47%的数据定位误差低于10m,计算时间少于0.3s。