基于多层感知器的气液两相流流型识别方法

通过对电阻层析成像数据采集原理和深度学习网络的研究,提出了一种基于阵列电阻值和多层感知器深度学习网络相结合的流型识别方法。利用电阻层析成像系统中的16个电极传感器来获取流型样本数据,并构建出流型识别数据库,然后对多层感知器深度学习网络进行训练,获得可以识别不同流型的网络。实验结果表明,采用阵列电阻值结合多层感知器网络对流型进行学习和识别的方法,流型识别准确率可以达到95%,解决了流型图像生成过程与数据特征预选过程中流型特征损失的问题,流型识别性能得到了提高。     

A benchmark data set for two-phase Coriolis metering

For more than a decade there has been growing interest in the use of Coriolis mass flow metering applied to two-phase (gas/liquid) and multiphase (oil/water/gas) conditions. It is well-established that the mass flow and density measurements generated from multiphase flows are subject to large errors, and a variety of physical models and correction techniques have been proposed to explain and/or to compensate for these errors. One difficulty is the absence of a common basis for comparing correction techniques, because different flowtube designs and configurations, as well as liquid and gas properties, may result in quite different error curves. Furthermore, some researchers with interests in the modelling aspects of the field may not have suitable multiphase laboratory facilities to generate their own data sets. This paper offers a small data set that may be used by researchers as a benchmark i.e. a common data set for comparing correction techniques. The data set was collected at the UK National Flow Laboratory TUV-NEL, using air and a viscous oil, and provides experimental points over a wide flow range (8:1 turndown) and with Gas Volume Fraction (GVF) values up to 60%. As a first investigation using the benchmark data set, we consider how data sparsity (i.e. the flow rate and GVF spacing in the experimental grid) affects the accuracy of a correction model. A range of neural network models are evaluated, based on different subsets of the benchmark data set. The data set and some exemplary code are provided with the paper. Additional data sets are available on a web site created to support this initiative.     

Ultrasensitive electrocatalytic detection of COX-2 rs20417: relying on 3D interconnected architecture of Pt-LSSUs@PAA nanostructures for sensor interface modification

Mutated COX-2 has become the new molecular marker of aspirin resistance. However, there is still a technical ‘bottleneck’ for direct and sensitive detection of circulating COX-2 mutant gene. In this work, we reported a simple and ultrasensitive electrochemical method for COX-2–765G/C (rs20417) detection for the first time. Polyallylamine (PAA) functionalised Pt nanostructures with long-spined sea urchin-like morphology (Pt-LSSUs@PAA) was synthesised by a simple chemical method for the construction of nano-sensing interface. Ru(NH₃)₆³⁺ is used as a primary electron acceptor that is electrostatically attracted to peptide nucleic acid modified electrodes and Fe(CN)₆^3? is introduced into the redox system as secondary electron acceptor to regenerate Ru³⁺ after electrochemical reduction for multiple redox cycles. Different pulse voltammetry was applied to record the electrochemical signals. Under optimal conditions, the DNA sensors showed a wide linear relationship, from 10?fM to 1?nM, with detection limits of 3.3?fM (S/N?=?3). This study will provide the basis for the precise use of aspirin, and it has important guiding value for individual drug testing of cardiovascular disease.     

基于模糊谱聚类的不确定蛋白质相互作用网络功能模块挖掘

针对谱聚类融合模糊C-means（FCM）聚类的蛋白质相互作用（PPI）网络功能模块挖掘方法准确率不高、执行效率较低和易受假阳性影响的问题，提出一种基于模糊谱聚类的不确定PPI网络功能模块挖掘（FSC-FM）方法。首先，构建一个不确定PPI网络模型，使用边聚集系数给每一条蛋白质交互作用赋予一个存在概率测度，克服假阳性对实验结果的影响；第二，利用基于边聚集系数流行距离（FEC）策略改进谱聚类中的相似度计算，解决谱聚类算法对尺度参数敏感的问题，进而利用谱聚类算法对不确定PPI网络数据进行预处理，降低数据的维数，提高聚类的准确率；第三，设计基于密度的概率中心选取策略（DPCS）解决模糊C-means算法对初始聚类中心和聚类数目敏感的问题，并对预处理后的PPI数据进行FCM聚类，提高聚类的执行效率以及灵敏度；最后，采用改进的边期望稠密度（EED）对挖掘出的蛋白质功能模块进行过滤。在酵母菌DIP数据集上运行各个算法可知，FSC-FM与基于不确定图模型的检测蛋白质复合物（DCU）算法相比，F-measure值提高了27.92%，执行效率提高了27.92%；与在动态蛋白质相互作用网络中识别复合物的方法（CDUN）、演化算法（EA）、医学基因或蛋白质预测算法（MGPPA）相比也有更高的F-measure值和执行效率。实验结果表明，在不确定PPI网络中，FSC-FM适合用于功能模块的挖掘。     

面向数据流结构的指令映射优化方法

在高性能计算领域,数据流是一类重要的计算结构,也在很多实际场景表现出很好的性能和适用性。在数据流计算模式中,程序是以数据流图来表示的,数据流计算中一个关键的问题是如何将数据流图映射到多个执行单元上。通过分析现有数据流结构的指令映射方法及其不足,提出了基于数据流结构的新型指令映射优化方法。主要是根据多地址共享数据包的特性对指令映射方法进行优化,延迟多地址共享数据路由包的拆分,减少网络拥堵。     

Simultaneous identification of structural stiffness and mass parameters based on Bare-bones Gaussian Tree Seeds Algorithm using time-domain data

This paper mainly illustrates the Tree Seeds Algorithm (TSA) to tackle structural damage identification problem. The damage model is simulated by the alterations of both stiffness and mass parameters. The objective function is introduced by minimizing the differences between the measured and calculated acceleration data. To enhance the performance of the standard TSA, two modifications including the bare-bones Gaussian updated mechanism and the withering process are introduced. The modified algorithm is named after the BGTSA. In the numerical simulation part, the BGTSA is firstly used to make comparisons with several state-of-the-art algorithms on the CEC05. Secondly, the BGTSA is utilized to deal with the structural damage identification problem by optimizing the acceleration-based nonlinear objective function. Numerical experiments involving a simply supported beam and a truss are carried out to verify the effectiveness of the proposed algorithm. The final results show that with low amount of acceleration data, the BGTSA can acquire better identification results compared with other evolutionary algorithms. Therefore the proposed algorithm could be viewed as a potential tool to solve the structural damage identification problem.     

面向多源数据融合的稀疏表示目标跟踪

传统的基于稀疏表示的目标跟踪方法主要利用目标的灰度特征构建稀疏表示模型。由于灰度特征对光照变化敏感，这会影响目标跟踪在复杂场景下的鲁棒性。基于多源数据融合的目标跟踪可以明显提升目标跟踪鲁棒性，但如何有效融合不同维度，不同类型的多源目标特征成为基于多源数据融合的目标跟踪所要解决的关键问题。提出了一个基于目标状态以及灰度特征的稀疏表示目标跟踪方法。所提出的方法可通过基于核函数表示的稀疏表示模型，在探究目标状态以及灰度特征相关性的基础上，将两种不同维度的特征进行有效融合，提升目标跟踪在复杂场景下的鲁棒性。     

基于聚类的数据加权优化在犯罪预测中的应用

近年来，我国传统暴力犯罪与成年人犯罪呈下降态势，但是，犯罪案由层出不穷。为有效提升公安实践工作中犯罪预测能力，打击各类违法犯罪事件,本文针对犯罪数据，提出一种新型犯罪预测模型。利用密度聚类分析方法将犯罪数据分类，然后进行数据降维提取关键属性生成特征数据，继而对特征数据进行加权优化并采用机器学习的方式对特征数据进行学习，从而预测犯罪案由。实验结果表明，与传统方法相比，本文方法具有更好的预测效果，为公安实践工作中类似案件的侦破和预防，提供新的路径支撑。