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.     

电力系统自动化发展方向分析

计算机技术、控制技术及信息技术的发展,电力系统自动化面临着空前的变革,多媒体技术、智能控制技术将迅速进入电力系统自动化领域。     

A Novel Beam Search to Improve Neural Machine Translation for English-Chinese

Neural Machine Translation (NMT) is an end-to-end learning approach for automated translation, overcoming the weaknesses of conventional phrase-based translation systems. Although NMT based systems have gained their popularity in commercial translation applications, there is still plenty of room for improvement. Being the most popular search algorithm in NMT, beam search is vital to the translation result. However, traditional beam search can produce duplicate or missing translation due to its target sequence selection strategy. Aiming to alleviate this problem, this paper proposed neural machine translation improvements based on a novel beam search evaluation function. And we use reinforcement learning to train a translation evaluation system to select better candidate words for generating translations. In the experiments, we conducted extensive experiments to evaluate our methods. CASIA corpus and the 1,000,000 pairs of bilingual corpora of NiuTrans are used in our experiments. The experiment results prove that the proposed methods can effectively improve the English to Chinese translation quality.     

基于深度语义分割的多源遥感图像海面溢油监测

针对遥感图像海面溢油区域通常受到斑噪声以及强度不均等因素的影响,从而导致溢油区域监测效果较差的问题,本文引入了深度语义分割的方法,将深度卷积神经网络与全连接条件随机场相结合,形成端对端连接。以Resnet结构为基础,首先通过深度卷积神经网络对多源遥感图像粗分割并作为输入,然后经过改进的全连接条件随机场,利用高斯成对势和平均场近似定理,建立条件随机场形成递归神经网络作为输出。通过多源遥感图像对海面溢油区域进行监测,并利用可见光图像估计溢油区域面积。实验在所建立的多源遥感图像数据集上与其它先进模型进行对比,结果表明本文方法提高了溢油区域的分割精度以及精细细节程度,平均交并比为82.1%,监测效果具有明显地改善。     

Local weather prediction system for a heating plant using cognitive approaches

Present-day requirements emphasize the need of saving energy. It relates mainly to industrial companies, where the minimization of energy consumption is one of their most important tasks they face. In our paper, we deal with the design of the so-called weather prediction system (WPS) for the needs of a heating plant. The primary task of such a WPS is timely predicting expected heat consumption to prepare the technology characterized by long delays in advance. Heat prediction depends primarily on weather so the crucial part of WPS is the weather, especially temperature, prediction. However, a prediction system needs a variety of further data, too. Therefore, WPS must be regarded as a complex system, including data collection, its processing, own prediction and eventual decision support. This paper gives the overview about existing data processing systems and prediction methods and then it describes a concrete design of a WPS with distributed data measuring points (stations), which are processed using a structure of neural networks based on multilayer perceptrons (MLP) with a combination of fuzzy logic. Based on real experiments we show that also such simple means as MLPs are able to solve complex problems. The paper contains a basic methodology for designing similar WPS, too.     

A neural network based technique for vibration characterization using Gaussian laser beams

This paper presents a neural network technique combined with an optical measurement system for the characterization of mechanical vibrations in industrial machinery. In the proposed system, the Gaussian beam of a laser source illuminates on an array of photodetectors. If either the laser source or the photodetector array is coupled with a vibrating system, then the optical powers detected by the photodetectors will vary accordingly, and are expected to reflect the magnitude and frequency of the X–Y planar vibrations of the monitored system. The time-varying optical powers are input to an artificial neural network-based vibration monitoring system which maps the power distributions to the X–Y position of the laser beam center. An experimental setup of the system is built and used for training and testing purposes. The obtained experimental results demonstrate the adequacy of combining optical techniques with neural networks to estimate the vibration frequency and magnitude. Estimated frequencies were within 1% of the actual ones, and the estimated magnitudes were within 29% of the actual magnitudes when using a chirp signal in the training phase. The magnitude estimation percentage error was further reduced below 12% when the neural network was trained with a decaying chirp signal.     

Root cause analysis of process fault conditions on an industrial concentrator circuit by use of causality maps and extreme learning machines

Model-based fault diagnosis tends to be too expensive or time-consuming to apply in the mineral processing industries, owing to the complexity and variability of operations. In contrast, data-based methods are inexpensive, but do not exploit the availability of first principle knowledge of plant operations. In this investigation, the use of process causality maps in conjunction with data-based fault diagnosis is considered as a hybrid methodology that can leverage the advantages of both approaches. Extreme learning machine algorithms are used to implement the data-based component of the approach. These algorithms can be deployed rapidly on large-scale systems and have the ability to deal with highly nonlinear systems. Two different variants are considered, viz. one used in combination with principal component analysis, as well as one with a bagging algorithm for fault diagnosis and applied to an industrial concentrator circuit in South Africa. The use of process causality maps led to significantly more effective fault diagnosis, while the use of extreme learning machines in combination with principal component analysis likewise allowed markedly better fault detection and diagnosis. In contrast, fault diagnosis with the bagging approach did not perform particularly well, owing to the high degree of correlation between the variables, which made it difficult to isolate individual causal variables.     

基于改进卷积神经网络的单幅图像超分辨率重建方法

对于重建图像存在的边缘失真和纹理细节信息模糊的问题，提出一种基于改进卷积神经网络（CNN）的图像超分辨率重建方法。首先在底层特征提取层以三种插值方法和五种锐化方法进行多种预处理操作，并将只进行一次插值操作的图像和先进行一次插值后进行一次锐化的图像合并排列成三维矩阵；然后在非线性映射层将预处理后构成的三维特征映射作为深层残差网络的多通道输入，以获取更深层次的纹理细节信息；最后在重建层为减少图像重建时间在网络结构中引入亚像素卷积来完成图像重建操作。在多个常用数据集上的实验结果表明，与经典方法相比，所提方法重建图像的纹理细节信息和高频信息能得到更好的恢复，峰值信噪比（PSNR）平均增加0.23 dB，结构相似性（SSIM）平均增加0.0066。在保证图像重建时间的前提下，所提方法更好地保持重建图像的纹理细节并减少图像边缘失真，提升重建图像的性能。     

基于深度神经网络的表面划痕识别方法

为实现亮度不均的复杂纹理背景下表面划痕的鲁棒、精确、实时识别，提出一种基于深度神经网络的表面划痕识别方法。用于表面划痕识别的深度神经网络由风格迁移网络和聚焦卷积神经网络（CNN）构成，其中风格迁移网络针对亮度不均的复杂背景下的表面划痕进行预处理，风格迁移网络包括前馈转换网络和损失网络，首先通过损失网络提取亮度均匀模板的风格特征和检测图像的知觉特征，对前馈转换网络进行离线训练，获取网络最优参数值，最终使风格迁移网络生成亮度均匀且风格一致的图像，然后，利用所提出的基于聚焦结构的聚焦卷积神经网络对生成图像中的划痕特征进行提取并识别。以光照变化的金属表面为例，进行划痕识别实验，实验结果表明：与需要人工设计特征的传统图像处理方法及传统深度卷积神经网络相比，划痕漏报率低至8.54%，并且收敛速度更快，收敛曲线更加平滑，在不同的深度模型下均可取得较好的检测效果，准确率提升2%左右。风格迁移网络能够保留完整划痕特征的同时有效解决亮度不均的问题，从而提高划痕识别精度；同时聚焦卷积神经网络能够实现对划痕的鲁棒、精确、实时识别，大幅度降低划痕漏报率和误报率。     

基于孪生神经网络的时间序列相似性度量

在时间序列分类等数据挖掘工作中，不同数据集基于类别的相似性表现有明显不同，因此一个合理有效的相似性度量对数据挖掘非常关键。传统的欧氏距离、余弦距离和动态时间弯曲等方法仅针对数据自身进行相似度公式计算，忽略了不同数据集所包含的知识标注对于相似性度量的影响。为了解决这一问题，提出基于孪生神经网络（SNN）的时间序列相似性度量学习方法。该方法从样例标签的监督信息中学习数据之间的邻域关系，建立时间序列之间的高效距离度量。在UCR提供的时间序列数据集上进行的相似性度量和验证性分类实验的结果表明，与ED/DTW-1NN相比SNN在分类质量总体上有明显的提升。虽然基于动态时间弯曲（DTW）的1近邻（1NN）分类方法在部分数据上表现优于基于SNN的1NN分类方法，但在分类过程的相似度计算复杂度和速度上SNN优于DTW。可见所提方法能明显提高分类数据集相似性的度量效率，在高维、复杂的时间序列的数据分类上有不错的表现。