基于 CEEMDAN 和概率神经网络的 起伏振动气液两相流型识别

起伏振动气液两相流型准确识别对漂浮核动力平台安全稳定运行有重要意义。通过对比静止和起伏振动管道的压差信号以及对应的频谱图发现,起伏振动管道内的压差信号波动幅度更大且包含更多的频率分量,两种流型均含有主频率,该频率为起伏振动频率。针对起伏振动状态气液两相流压差信号的复杂性,分别采用自适应白噪声的完备总体经验模态分解(CEEMDAN)和集合经验模态分解(EEMD)对小波降噪后的压差信号进行模式分解,发现CEEMDAN能够在减少模式分量的同时获得更多有效的分量。通过计算spearman相关系数选择具有表征意义的IMF分量进行Hilbert变换计算能量作为特征值,采用概率神经网络对流型进行识别。结果表明,采用CEEMDAN进行模式分解结合概率神经网络的识别方法准确率达到95.83%,能够用于起伏振动下气液两相流型识别。     

基于希尔伯特-黄变换和隐马尔可夫模型的气液两相流流型识别方法

为了研究垂直上升管中的气液两相流的流型,利用自制的多电导探针的测量系统采集了四种典型流型的电导波动信息.由于气液两相流电导波动信号的非平稳特征以及神经网络学习收敛慢等问题,提出了一种基于希尔伯特-黄变换(hilbert-huang transform,HHT)和隐马尔可夫模型(hidden markov model,HMM)的两相流流型识别方法.该方法首先将信号经验模态分解(empirical mode decomposition,EMD)后的固有模态函数(IMFs)进行希尔伯特变换得到其幅值能量,并将其作为特征向量,输入到已经训练完毕的各状态HMM中,实现了对气液两相流的流型识别.实验结果表明:该方法能很好的识别垂直管内的4种流型,而且优于BP神经网络,从而为流型识别开辟了一条新的途径.     

Hilbert-Huang变换及其在纱线故障诊断中的应用

阐述了Hilbert-Huang 变换的基本原理,以实际的纱线信号为例,通过经验模态分解(EMD)将纱线信号分解成有限个固有模态函数(IMF),并求得纱线信号的Hilbert谱.结合分析纱线信号的波谱图和Hilbert谱图,判断出故障产生的原因.     

气-液两相流压差波动信号中噪声的辨识

压差波动信号由于其容易测量,而且包含流动的很多信息而被广泛用于气液两相流的研究中。在测量时由于难免受到各种干扰,这些干扰会影响非线性分析的精度和结果。因此,在利用压差波动信号进行非线性分析之前,通常要对其进行滤波处理。究竟多高频率以上的压差波动信号才是噪声,并如何对其进行定量检测的研究就非常重要。这里对水平管内气液两相流压差波动信号用小波包进行分解,得到按16Hz等间隔分段的16个不同频带信号,用自相关函数研究了各频段信号的线性相关性。通过理论证实了气液两相流压差波动信号中,64Hz以上信号不存在线性相关性,可以认为其是噪声信号。研究指出通过气液两相流压差波动信号来研究流型信息时,应考虑64Hz以下信号。     

基于经验模态分解的瞬时相位分析方法的应用

提出了基于经验模态分解的瞬时相位分析的新方法。通过对振动信号作经验模态分解得到信号的固有模态函数，再求出各个固有模态函数的Hilbert变换，得到信号的瞬时相位．通过瞬时相位的傅里叶分析就可提取信号特征。介绍了该方法的基本原理，并应用于齿轮箱轴承的故障诊断研究，通过选取表征轴承故障的固有模态函数进行瞬时相位和傅里叶分析，就可提取轴承故障振动信号的特征。通过对轴承故障实验信号的分析．表明该方法能有效地诊断轴承的故障。     

小波变换辨识流型的一种新方法研究

本文提出了一种基于小波分析进行水平管气液两相流流型辨识的新方法。该方法以测试管段的管程压降△P的波动作为测量信号,通过对其小波变换结果的分析,根据不同流型能量分布的定量指标确定了流型判别规则。实验结果表明,这种判别方法能有效地实现对水平管层状流、波状流、塞状流、弹状流等四种典型流型的在线辨识。     

基于经验模态分解和BP神经网络的油气两相流流型辨识

基于经验模态分解（empidcal mode decomposition，EMD）BP神经网络，提出了油气两相流流型辨识的新方法。应用EMD将差压信号分解成不同频率尺度上的单组分之和，并提取组分的归一化能量作为流型辨识特征量。BP神经网络以这些能量特征量为输入对油气两相流不同流型（包括泡状流、塞状流、层状流、弹状流和环状流）进行分类。实验结果表明，本文提出的流型辨识方法是有效的，其中泡状流、塞状流、层状流、弹状流和环状流的辨识精度分别为100％、89．4％,93．3％、96.3％和96.9％。     

气液两相流系统的小波软测量技术

提出了一种基于小波分析技术的气液两相流流型判别软测量模型.气液两相流中简单、易测而且可靠的差压波动信号作为软测量模型中的辅助变量,气液两相流的流型作为模型中的主导变量,利用小波分析技术建立了主导变量和辅助变量之间的数学关系,进而通过测量差压波动信号可以对气液两相流的流型进行判别.实验证明所建立的软测量模型是有效的.     

不波变换辨识流形的一种新方法研究

本文提出了一种基于小波分析进行水平管气液两相流流型辨识的新方法。该方法以测试管段的管程压降△Ｐ的波动作为测量信号,通过对其小波变换结果的分析,根据不同流型能量分布的定量指标确定了流型判别规划。实验结果表明,这种判别方法能有效地实现对水平管层状流、波状流、塞状流、弹状流等四种典型流形的在线辨识。     

应用Hilbert-Huang变换的齿轮箱故障诊断

提出了基于经验模态分解(EMD)的齿轮箱故障诊断HHT方法,介绍了Hilbert-Huang变换理论及其算法.此后以1台现场轧机故障减速机为对象,对采集的故障信号进行EMD分解,得到固有模态函数(IMF)分量,然后对所有固有模态函数进行Hilbert变换处理,所得三维图和边际谱图较为清晰地表达了故障信息,说明了该方法在工程应用中的适用性.     

Response of a dual triangulate bluff body vortex flowmeter to oscillatory flow

In this paper, on an experimental facility, the measurement characteristics of a diameter 50 mm dual triangulate bluff body vortex flowmeter in steady flow and oscillatory flow were investigated. Then, the Hilbert Huang Transformation (HHT) method was used to assess the anti-interference performances and the vortex street stability in oscillatory flow for the dual triangulate bluff body vortex flowmeter and a single bluff body vortex flowmeter. Offline simulation was carried out on the anti-interference performances of the dual triangulate bluff body vortex flowmeter signal noise in oscillatory flow by the method of the EMD-scales filter. The major findings are: (a) in most case, the EMD-scales filter may be as good at de-noising effect for the dual bluff body vortex flowmeter in oscillatory flow than that for the single bluff body vortex flowmeter in oscillatory flow. The vortex street stability in oscillatory flow for the dual bluff body is similar to that for the single bluff body. (b) In some special case, the EMD-scales filter is unable to play a better de-nosing role for the dual bluff body vortex flowmeter in oscillatory flow. The invalid condition of the EMD-scales filter for the dual bluff body vortex flowmeter in oscillatory flow is different to that of the single bluff body vortex flowmeter and it was advanced in this paper. (C) The vortex street stability for the dual bluff body vortex flowmeter is better than that for the single bluff body vortex flowmeter.     

Research on signal amplitude of the Kármán vortex street in gas–liquid two-phase flow with high void fraction

Based on Biot–Savart law and single-phase flow Kármán vortex characteristics, flow field has been analyzed when gas–liquid flow past a fixed bluff body with high void fraction. Vortex signal characteristics have been studied for stratified two-phase flow on atmospheric conditions in a horizontal pipe. To discuss the relation between void fraction and vortex signal amplitude spectrum, this paper sets up the vortex-induced pressure field model for gas–liquid two-phase flow and gives the relationship between void fraction and relative amplitude spectrum of two-phase flow to single-phase flow. An algorithm is proposed for predicting the two-phase flow parameters. Experiments were performed using air–water as working fluid along with a test tube diameter of 50 mm, at gas volume flow rate of 20–68 m³/h, and void fraction of 0.9–1. The results indicate that calculations by the vortex-induced pressure field model on the amplitude spectrum of vortex signal are in good agreement with the experimental data, and relative errors of the algorithm predictions on gas volume flow rate and liquid volume flow rate are 0.08 and 0.56, respectively.     

The mapping of vortex fields around single and dual bluff bodies

The development and use of a system to measure important parameters of the vortex field in a coordinate grid around several single bluff bodies and dual bluff body combinations are described. Two hot wire sensors were used: a reference sensor at a fixed position and a moving sensor with position coordinates adjusted by using computer control. The strength, frequency and regularity of the vortex shedding are found from the auto-power spectral density of the moving sensor signal. The phase difference between moving and reference sensor signals is found from the cross-power spectral density function between the signals. The results are presented as maps of vortex parameters plotted as a function of sensor position coordinates. The main features of these maps, including the enhancement of vortex shedding from certain dual bluff body combinations, is then discussed.     

基于多频带谱熵的水平气液两相流结构复杂性分析

针对传统功率谱熵只能刻画总体系统结构复杂性问题,提出既能从宏观角度又能从微观角度反映系统结构复杂性的多频带谱熵的分析方法。分析几种典型信号的多频带谱熵特征,验证多频带谱熵方法的可行性及抗干扰能力;利用环形电导传感器阵列获取动态实验测试数据,计算水平气液两相流波状流、塞状流和弹状流含水率波动的多频带谱熵值,分析3种流型结构复杂性随频带因子变化的演变规律。实验结果表明:波状流的多频带谱熵最低,弹状流多频带谱熵值最高;塞状流的多频带谱熵居于波状流与弹状流之间。3种流型在分析频率为0~8.3 Hz时,呈现近似线性变化的结构复杂性特征以及较强的流型区分度,可以作为流型识别的准则。     

Flow pattern identification and measurement techniques in gas-liquid-liquid three-phase flow: A review

The simultaneous flow of gas, oil, and water forms various flow patterns due to the complex interfacial relationships. Three-phase flow patterns are classified as the gas-liquid and liquid-liquid flow patterns. Pressure drop, void fraction, liquid holdup, and phase distribution are important characteristics of the three-phase flow. These characteristics are generally associated with the three-phase flow patterns. Hence, the knowledge about flow patterns can help to predict the overall behavior of the three-phase flow. Studies have been conducted to identify three-phase flow pattern and their characteristics at various superficial velocities of gas, oil, and water. The major purpose of the studies is to gather information about the three-phase co-current flow and use it for improvement of the efficiency of the flow systems. Therefore, the accuracy of the measurement technique is critical. Several types of flow pattern identification and measurement techniques have been developed to improve accuracy and provide high-quality results. In this article, classical and advanced techniques used for the three-phase flow identification and measurement have been reviewed. The survey will help the researchers working in the area of multiphase flow to choose the right technique based on the objectives of the studies.     

Doppler spectrum analysis and flow pattern identification of oil-water two-phase flow using dual-modality sensor

Horizontal oil-water two-phase flow widely exists in petroleum and chemical engineering industry, where the oil and water are usually transported together. As one of most importance process parameters to describe the two-phase flow, the flow pattern can reflect the flow characteristics of inner flow structure and phase distribution. The identification of flow pattern will contribute to develop more accurate measurement model for flow rate or phase fraction and ensure the safety and efficiency of operation in industry. A dual-modality sensor combining with continuous wave ultrasonic Doppler sensor (CWUD) and auxiliary conductance sensor, was proposed to identify flow patterns in horizontal oil-water two-phase flow. In particular, the oil-water flow characteristic was analyzed from Doppler spectrum based on the CWUD sensor. Besides, the dimensionless voltage parameter based on conductance sensor was applied to provide the information of continuous phase in the fluid. Several statistical features were directly extracted without any complicated processing algorithm from Doppler and conductance signals. The extracted features are put into a multi-classification Support Vector Machine (SVM) model to classify five oil-water flow patterns. The results show that the overall identification accuraccy of 94.74% is satisfactory for horizontal oil-water two-phase flow. It also demonstrates that the noninvasive ultrasonic Doppler technique not only can be used for flow velocity measurement but also for flow pattern identification.     

Horizontally installed cone differential pressure meter wet gas flow performance

Differential pressure (DP) meters which utilise a cone as the system’s primary element are increasingly being used to measure wet natural gas flows (i.e. mixtures of natural gas, light hydrocarbon liquids and water). It is therefore important to understand this meter’s response to wet natural gas flows. Research into the wet gas response of the horizontally installed cone DP meter is discussed in this paper. Consideration is given to the significant influence of the liquid properties on wet gas flow patterns and the corresponding influence of the flow pattern on the cone DP meter’s liquid phase induced gas flow rate prediction error. A wet natural gas flow correlation for 4 in. 0.75 beta ratio cone DP meters with natural gas, hydrocarbon liquid and water flow has been developed from multiple data sets from three different wet gas flow test facilities. This corrects the liquid induced gas flow rate prediction error of a wet gas flow up to a Lockhart–Martinelli parameter of 0.3, for a known liquid flow rate of any hydrocarbon liquid/water ratio, to ±4% at a 95% confidence level.     

Voidage measurement of gas–liquid two-phase flow based on Capacitively Coupled Contactless Conductivity Detection

Based on Capacitively Coupled Contactless Conductivity Detection (C⁴D) technique, a new method for the voidage measurement of conductive gas–liquid two-phase flow is proposed. 15 Conductance signals, which reflect voidage distribution of gas–liquid two-phase flow, are obtained by a six-electrode C⁴D sensor. With the conductance signals, the flow pattern of gas–liquid two-phase flow is identified by flow pattern classifiers and then the voidage measurement is implemented by a corresponding voidage measurement model (for each typical flow pattern, a corresponding voidage measurement model is developed). The conductance measurement of the six-electrode C⁴D sensor is implemented by phase sensitivity demodulation (PSD) method. The flow pattern classifiers and the voidage measurement models are developed by partial least squares (PLS) technique and least squares support vector machine (LS-SVM) technique. Static voidage measurement experiments and dynamic voidage measurement experiments show that the proposed voidage measurement method is effective, the developed six-electrode C⁴D sensor is successful and the measurement accuracy is satisfactory.     

IDENTIFICATION OF GAS–SOLID TWO-PHASE FLOW REGIMES USING HILBERT–HUANG TRANSFORM AND NEURAL-NETWORK TECHNIQUES

This work presents a new methodology for flow regime identification in a gas–solid two-phase flow system. The approach of identification employs the artificial neural network (ANN) technique, considering the applications with electrostatic sensor as a measuring device and Hilbert–Huang transformation (HHT) as the post-processing method. The electrostatic fluctuation signals detected from an electrostatic sensor are processed using HHT to gain the Hilbert marginal spectrums. Then four characteristic parameters of the marginal spectra are extracted as the input of BP neural network for flow regime identification. They are subband energy (SE), first-order difference of subband energy (DSE), subband energy cepstrum coefficients (SECC), and first-order difference of the subband energy cepstrum coefficients (DSECC). The results show that the characteristic parameters of the Hilbert marginal spectrum of the electrostatic signal can identify the three flow regimes of gas–solid two-phase flow in a horizontal pipe, especially the DSECC.