基于时间序列数据挖掘的旋转机械预诊断方法研究

基于状态的维护(CBM, Condition Based Maintenance)是近年来新兴的一种设 备维护策略，它的基本理念是在机械设备需要维护的时候才对其进行维护，强调 维护要及时、准确和经济。采用这种维护策略，能够提高工业生产的安全性和可 靠性，系统地降低企业运营成本。 机械设备状态预诊断是实现CBM 的核心支撑技术，对其进行深入研究，对推 动CBM 的发展具有重要意义。但是，由于相关研究起步不久，目前预诊断技术还 未能得到很好的实现，研究人员有必要不断地尝试各种新的有效方法来更好地解 决这一问题，加快其实现方法与技术应用的成熟进程。基于此，本文从数据挖掘 的角度，探索了机械设备预诊断新的解决方法和途径，深入研究和探讨了基于时 间序列数据挖掘的旋转机械预诊断方法。本文的主要工作包括： 1. 结合CBM 的基本理念和应用实际的需求，对机械设备状态预诊断的基本 内涵进行了系统分析。将状态评估、故障预测和剩余有效使用寿命预测三个预诊 断基本功能进一步抽象，提出了包含特征提取、状态预测和模式匹配三个子问题 的预诊断一般流程模式。在详细分析机械设备状态预诊断理论方法和应用技术研 究现状的基础上，提出了预诊断技术研究的发展趋势及各子问题的研究侧重点。 并对利用时间序列数据挖掘这一理论方法解决机械设备状态预诊断问题的可行性 进行了分析。 2. 针对具有波动频繁、噪声干扰严重等特点的原始振动量时间序列无法直接 用于旋转机械性能状态分析的问题，结合全息诊断信息融合分析旋转机械振动全 貌的思想，提出了全息状态矩阵的概念并给出定义，用类时间轴上的多维序列表 征转子系统振动全貌，以实现振动量时间序列的高级表示，为后续预测与匹配分 类工作提供良好的数据源，同时增强全息诊断的信息检索和知识自动获取的能力。 3. 将旋转机械性能状态预测，归结为旋转机械设备维护应用背景下的一维数 值型时间序列预测问题来进行深入研究。针对现有预测方法长期预测能力较弱， 且自动化水平低的不足，提出了用于旋转机械性能状态预测的ARIMA 动态间隔预 测法。该方法以动态间隔获取时间序列样本建模并预测的策略，提高了ARIMA 模 型用于设备状态长期预测的准确性，并且能够实现建模与预测的自动化，满足CBM 系统的实时性要求。 4. 针对全息状态矩阵表示的旋转机械性能状态特征数据，提出了一种全息状 态矩阵相似性匹配方法。结合旋转机械预诊断领域应用的特点定义了全息状态矩 阵的相似性度量模型，基于全息状态矩阵近似距离三角不等式设计了剪枝搜索策 略，并在此基础上设计了全息状态矩阵相似性高效准确匹配算法，不需要借助专家经验和人工识别确认，在一定阈值范围内能够实现高质量的旋转机械性能状态 相似性匹配。 5. 旋转机械基本振动量特征时间序列具有海量、超高维度、短期波动频繁和 大量噪声等特征，与时间序列数据挖掘传统应用的金融商业领域数据不同，直接 采用传统方法会存在搜索速度大幅度降低的问题。针对这一问题，提出了基于随 机投影的时间序列相似性搜索方法。该方法利用近年来新兴的随机投影统计学降 维法，将原始时间序列集映射到低维空间，并利用R*树进行索引，能够在保持高 准确率的同时，实现旋转机械基本振动量特征时间序列相似性快速搜索。 6. 针对现有机械设备性能状态分类方法不考虑误分类代价的问题，提出了一 种代价敏感直推式旋转机械设备性能状态分类法。该方法将代价敏感分类和直推 式学习的基本思想和理论相结合，采用一种代价敏感的直推式分类机制，实现了 机械设备性能状态的代价敏感分类。该方法在保证较高分类准确率的基础上，明 显地降低了误分类总代价。 7. 基于CBM 的基本理念，设计了旋转机械CBM 系统的基本结构，并以本 文理论方法的研究成果为核心，详细设计了各模块的基本功能和处理逻辑，采用 VC#.net 与Matlab 混合编程的方式开发了一个面向大型旋转机械的CBM 系统原 型，以验证本文机械设备预诊断方法研究成果的可操作性和实用性，为CBM 系统 应用技术研究做出了有益的探索。

Effect of spontaneous and piezoelectric polarization on intersubband transition in AlxGa1-xN-GaN quantum well

A self-consistent solution of conduction band profile and subband energies for AlxGa1-xN-GaN quantum well is presented by solving the Schrodinger and Poisson equations. A new method is introduced to deal with the accumulation of the immobile charges at the AlxGa1-xN-GaN interface caused by spontaneous and piezoelectric polarization in the process of solving the Poisson equation. The effect of spontaneous and piezoelectric polarization is taken into account in the calculation. It also includes the effect of exchange-correlation to the one electron potential on the Coulomb interaction. Our analysis is based on the one electron effective-mass approximation and charge conservation condition. Based on this model, the electron wave functions and the conduction band structure are derived. We calculate the intersubband transition wavelength lambda(21) for different Al molar fraction of barrier and thickness of well. The calculated result can fit to the experimental data well. The dependence of the absorption coefficient a on the well width and the doping density is also investigated theoretically. (C) 2004 American Vacuum Society.

边坡动力响应分析及应用研究

As a marginal subject, dynamic responses of slopes is not only an important problem of engineering geology (Geotechnical problem), but also of other subjects such as seismology, geophysics, seismic engineering and engineering seismic and so on. Owning to the gulf between different subjects, it is arduous to study dynamic responses of slopes and the study is far from ripeness. Studying on the dynamic responses of slopes is very important in theories as well as practices. Supported by hundreds of bibliographies, this paper systemically details the development process of this subject, introduces main means to analyze this subject, and then gives brief remarks to each means respectively. Engineering geology qualitative analysis is the base of slopes dynamic responses study. Because of complexity of geological conditions, engineering geology qualitative analysis is very important in slopes stability study, especially to rock slopes with complex engineering geology conditions. Based on research fruits of forerunners, this paper summarizes factors influencing slopes dynamic stability into five aspects as geology background, stratums, rock mass structure, and topography as well as hydrogeology condition. Based on rock mass structure controlling theory, engineering geology model of the slope is grouped into two typical classes, one is model with obvious controlling discontinuities, which includes horizontal bedded slope, bedding slope, anti-dip slope, slide as well as slope with base rock and weathered crust; the other is model without obvious controlling discontinuities, which includes homogeneous soil slope, joint rock mass slope. Study on slope failure mechanism under dynamic force, the paper concludes that there are two effects will appear in slope during strong earthquake, one is earthquake inertia force, the other is ultra pore pressure buildup. The two effects lead to failure of the slope. To different types of slope failure, the intensity of two effects acting on the slope is different too. To plastic flow failure, pore pressure buildup is dominant; to falling rock failure and toppling failure, earthquake inertia force is dominant in general. This paper briefly introduces the principle of Lagrangian element method. Through a lot of numerical simulations with FLAC3D, the paper comprehensively studies dynamic responses of slopes, and finds that: if the slope is low, displacement, velocity and acceleration are linear enlarging with elevation increasing in vertical direction; if the slope is high enough, displacement, velocity and acceleration are not linear with elevation any more, on the other hand, they fluctuate with certain rhythm. At the same time, the rhythm appears in the horizontal direction in the certain area near surface of the slope. The distribution form of isoline of displacement, velocity and acceleration in the section of the slope is remarkably affected by the slope angle. In the certain area near the slope surface, isoline of displacement, velocity and acceleration is parallel to the surface of the slope, in the mean time, the strike direction of the extreraum area is parallel to the surface of the slope too. Beyond this area, the isoline direction and the strike direction of the extremum area turn to horizontal with invariable distance. But the rhythm appearing or not has nothing to with the slope angle. The paper defines the high slope effect and the low slope effect of slopes dynamic responses, discusses the threshold height H^t of the dynamic high slope effect, and finds that AW is proportional to square root of the dynamic elastic moduli El P , at the same time, it is proportional to period Tof the dynamic input. Thus, the discriminant of H^t is achieved. The discriminant can tell us that to a slope, if its height is larger than one fifth of the wavelength, its response regular will be the dynamic high slope effect; on the other hand, its response regular will be the dynamic low slope effect. Based on these, the discriminant of different slopes taking on same response under the same dynamic input is put forward in this paper. At the same time, the paper studies distribution law of the rhythm extremum point of displacement, velocity and acceleration, and finds that there exists relationship of N = int among the slope height H, the number of the rhythm extremum VHlhro) point N and ffthre- Furthermore, the paper points out that if N^l, the response of the slope will be dynamic high slope effect; \fN

Clemaps: Multiple Alignment Of Protein Structures Based On Conformational Letters

CLEMAPS is a tool for multiple alignment of protein structures. It distinguishes itself from other existing algorithms for multiple structure alignment by the use of conformational letters, which are discretized states of 3D segmental structural states. A letter corresponds to a cluster of combinations of three angles formed by C-alpha pseudobonds of four contiguous residues. A substitution matrix called CLESUM is available to measure the similarity between any two such letters. The input 3D structures are first converted to sequences of conformational letters. Each string of a fixed length is then taken as the center seed to search other sequences for neighbors of the seed, which are strings similar to the seed. A seed and its neighbors form a center-star, which corresponds to a fragment set of local structural similarity shared by many proteins. The detection of center-stars using CLESUM is extremely efficient. Local similarity is a necessary, but insufficient, condition for structural alignment. Once center-stars are found, the spatial consistency between any two stars are examined to find consistent star duads using atomic coordinates. Consistent duads are later joined to create a core for multiple alignment, which is further polished to produce the final alignment. The utility of CLEMAPS is tested on various protein structure ensembles.

Damage evaluation and damage localization of rock

Knowledge of damage accumulation and corresponding failure evolution are prerequisite for effective maintenance of civil engineering so as to avoid disaster. Based on statistical mesoscopic damage mechanics, it was revealed that there are three stages in the process of deformation, damage and failure of multiscale heterogeneous elastic-brittle medium. These are uniformly distributed damage, localized damage and catastrophic failure. In order to identify the transitions from scattering damage to macroscopically localized one, a condition for damage localization was given. The experiments of rock under uniaxial compression with the aid of observations of acoustic emission and speckle correlation do support the concept of localization. This provides a potential approach to properly evaluate damage accumulation in practice. In addition, it is found in the experiments that catastrophic failure displays critical sensitivity. This gives a helpful clue to the prediction of catastrophic failure. (C) 2004 Elsevier Ltd. All rights reserved.

Flow similariy applied in convection-based tilt sensor

The environment temperature has inevitable effects on property of the convect ion-based tilt sensors. It not only redefines the application, but also prevents the improvement of the sensor performance. Numerical simulation of the fluid flow in the chamber of a sensor was performed and the influence of the environment temperature was studied in this paper. At zero tilt angle, the temperature distribution along the perpendicular line cross the heat source at various environment temperatures was presented. It was found that the flow varied dramatically at different environment temperatures, which would cause the output signal vary accordingly, even when the tilt angle was kept at a constant, because this device works by sensing the change of flow. At the same condition, we present the numerical results when the temperature difference across the heat source and the environment was kept at the same, in those results, it was found that the temperature difference at every point along the perpendicular line cross the heat source keep the same, this result confirms the similarity principle of nature convection. Second, A method of eliminating environment temperature infect on property of convect ion-based tilt sensor, which is based on the theory of flow similarity, is proposed. It was found that a thermal transistance can be piped on the circuit of heat source to compensate the temperature of the heat source. A compensative circuit was specially designed which can keep flow similarity by changing heat source temperature in order to eliminate the influence of environment temperature. The experiment results show that above 70% temperature drift can be eliminated by this compensative circuit.

Study of noncollinear quasi-phase matching optical parametric amplification with group-velocity matching based on periodically poled crystals

The properties of noncollinear optical parametric amplification (NOPA) based on quasi-phase matching of periodically poled crystals are investigated, under the condition that the group velocity matching (GVM) of the signal and idler pulses is satisfied. Our study focuses on the dependence of the gain spectrum upon the noncollinear angle, crystal temperature, and crystal angle with periodically poled KTiOPO4 (PPKTP), periodically poled LiNbO3 (PPLN), and periodically poled LiTaO3 (PPLT), and the NOPA gain properties of the three crystals are compared. Broad gain bandwidth exists above 85 nm at a signal wavelength of 800 nm with a 532 nm pump pulse, with proper noncollinear angle and grating period at a fixed temperature for GVM. Deviation from the group-velocity-matched noncollinear angle can be compensated by accurately tuning the crystal angle or temperature with a fixed grating period for phase matching. Moreover, there is a large capability of crystal angle tuning.

Multi-channel 2-D pattern transfer model based on imaging optics for submarine laser uplink communication

An optical communication scheme of 2-D pattern transfer based on imaging optics for submarine laser uplink communication (SLUC) is suggested. Unlike the methods aiming at avoiding neighboring crosstalk used in traditional multi-channel optical beam transferring, we make full use of the overlapping of each spreading beam other than controlling divergence effect of each beam to avoid interference noise. The apparent parameters have been introduced to simplify theoretical analysis of optical pattern transfer problem involving underwater condition, with the help of which the complex beam propagation inside two kinds of mediums can be easily reduced to brief beam transfer only inside air medium. In this paper, optical transmission path and receiver terminal optics geometry have been described in detail. The link range equation and system uplink performance analysis have also been given. At last, results of a proof-of-concept experiment indicate good feasibility of the proposed SLUC model. © 2007 Elsevier GmbH. All rights reserved.

Growth and Characterization of GaSb-Based Type-II InAs/GaSb Superlattice Photodiodes for Mid-Infrared Detection

InAs/GaSb superlattice (SL) midwave infrared photovoltaic detectors are grown by molecular beam epitaxy on GaSb(001) residual p-type substrates. A thick GaSb layer is grown under the optimized growth condition as a buffer layer. The detectors containing a 320-period 8ML/8ML InAs/GaSb SL active layer are fabricated with a series pixel area using anode sulfide passivation. Corresponding to 50% cutoff wavelengths of 5.0 mu m at 77 K, the peak directivity of the detectors is 1.6 x 10(10) cm.Hz(1/2) W-1 at 77 K.

Spin current diode based on an electron waveguide with spin-orbit interaction

We propose a spin current diode which can work even in a small applied bias condition (the linear-response regime). The prototypal device consists of a hornlike electron waveguide with Rashba spin-orbit interaction, which is connected to two leads with different widths. It is demonstrated that when electrons are incident from the narrow lead, the generated spin conductance fluctuates around a constant value in a wide range of incident energy. When the transport direction is reversed, the spin conductance is suppressed strongly. Such a remarkable difference arises from spin-flipped transitions caused by the spin-orbit interaction. (c) 2008 American Institute of Physics.

A Cancer Recognition Method Based on DNA Microarray

The accurate cancer classification is of great importance in clinical treatment. Recently, the DNA microarray technology provides a promising approach to the diagnosis and prognosis of cancer types. However, it has no perfect method for the multiclass classification problem. The difficulty lies in the fact that the data are of high dimensionality with small sample size. This paper proposed an automatic classification method of multiclass cancers based on Biomimetic pattern recognition (BPR). To the public GCM data set, the average correct classification rate reaches 80% under the condition that the correct rejection rate is 81%.

Wavelength conversion based on degenerate-four-wave-mixing with continuous-wave pumping in silicon nanowire waveguide

We present a comprehensive numerical study on the all-optical wavelength conversion based on the degenerate four-wave-mixing with continuous-wave pumping in the silicon nanowire waveguide. It is well known that the conversion efficiency and the 3-dB bandwidth can be greatly affected by the phase-matching condition. Through proper design of the waveguide cross-section, its dispersion property can be adjusted to satisfy the phase-matching condition and therefore effective wavelength conversion can be achieved in a large wavelength range. Generally, the group velocity dispersion plays a dominant role in the wavelength conversion. However, the fourth-order dispersion takes an important effect on the wavelength conversion when the group velocity dispersion is near the zero-point. Furthermore, the conversion efficiency and the 3-dB bandwidth can also be affected by the interactive length and the initial pump power. Through the numerical simulation, the optimal values for the interactive length and the initial pump power, which are functions of the propagation loss, are obtained to realize the maximum conversion efficiency. (C) 2008 Elsevier B.V. All rights reserved.

Design and Fabrication of a Photonic Crystal Channel Drop Filter Based on an Asymmetric Silicon-on-Insulator Slab

We report on the design and fabrication of a photonic crystal (PC) channel drop filter based on an asymmetric silicon-on-insulator (SOI) slab. The filter is composed of two symmetric stick-shape micro-cavities between two single-line-defect (W1) waveguides in a triangular lattice, and the phase matching condition for the filter to improve the drop efficiency is satisfied by modifying the positions and radii of the air holes around the micro-cavities. A sample is then fabricated by using electron beam lithography (EBL) and inductively coupled plasma (ICP) etching processes. The measured 0 factor of the filter is about 1140, and the drop efficiency is estimated to be 73% +/- 5% by fitting the transmission spectrum.

Single-mode condition for silicon rib waveguides with large cross sections

The single-mode condition for silicon rib waveguides with large cross sections is obtained using a numerical method based on the finite-difference beam propagation method (FD-BPM). An ultrawide computation window is used to contain the wide mode profile near cutoff. Comparison with previous results shows that the formula predicted by the mode-matching technique is in a better agreement with our results. (C) 2004 Society of Photo-Optical Instrumentation Engineers.

Single-mode behaviour judgment of optical waveguides by imaginary-distance beam propagation method under perfectly matched layer boundary condition

Imaginary-distance beam propagation method under the perfectly matched layer boundary condition is applied to judge single-mode behaviour of optical waveguides, for the first time to our knowledge. A new kind of silicon-on-insulator-based rib structures with half-circle cross-section is presented. The single-mode behaviour of this kind of waveguide with radius 2mum is investigated by this method. It is single-mode when the slab height is not smaller than the radius.