样本分布

This paper is expanded into three main parts to study and analyze the stock time series.1 Through the normality tests, All the sample stock-data curves show different phenomenon of Aiguille and Fat-tail, as well as Left-Deviation, which proves that none of the sample series belong to normal distribution, in other words, the linear method can't be applied into the analysis process, so we adopted the R/S method instead.2 The application of R/S method has verified the long-term memory characteristics of the sample series, and Support Vector Regression is combined to enhance the precision of Average circle period.

文中对股票时间序列从以下三个方面进行了研究和分析：1通过正态性检验，发现各样本股票数据均表现出不同程度的尖峰和胖尾以及左偏现象，证明了所有的样本序列全都不属于正态分布，也就是说明它们不能够适用线性方法进行分析，必须采用重标极差分析方法的非线性方法。2应用R/S分析方法验证他们具有长期记忆的特性，并且结合支持向量回归机提高了测定平均循环周期时的精确性。

Real AdaBoost algorithm demands division of the sample space. The traditional finite division can not reflect the distribution of positive and negative samples.

连续AdaBoost算法要求对样本空间进行划分，传统的等距划分无法体现正负样本各自的分布规律。

This dissertation is concerned with some problems regarding the asymptotic property of nearest-neighbor density estimator for α—mixing and nagetive quadrant dependent samples. The results extend the corresponding theorems in i.i.d.

本文主要考虑了最近邻密度估计在α—混合序列和两两NQD混合序列情形下的大样本性质，推广了独立同分布和其它相依情形下最近邻密度估计的大样本性质，如相合性、渐近正态性等。

The paper consists of three parts. First, we give the calculation for every order central moment for the Logistic population, mean vector and covariance matrix of the order statistic of random sample from the Logistic population, and means of several special functions for the order statistic of random sample from the Logistic population. Secondly, we discuss the parameter estimation for the Logistic population. Finally, we solve the question about the goodness-of -fit test for the Logistic population.

研究的主要内容分三部分：第一部分是关于Logistic总体各阶中心矩、样本次序统计量期望向量与协方差阵及若干个特殊的样本次序统计量函数期望的计算问题；第二部分是关于Logistic总体分布参数的估计问题；第三部分是关于Logistic总体的拟合优度检验问题。

The effect of tribology factors on bearing vibration is studied experimentally. AR model and WPT are used to processing the normal arid abnormal signals. The results are as follows:1. The vibration of ball bearings is caused by the tribology action of contact pairs of ball and ring races and has the essential of tribo-dynamics. Any factors affecting the tribology characteristics of contact pairs will affect vibration and noise of bearings consequently.2. Vibration characteristics of bearing keep almostly unchanged with the increasing of axial load expecting the increasing of nature frequencies. When radial load increasing with axial load unchanged, nature frequencies keep unchanged but some new vibration peaks appears in the spectrum.3. Vibration amplitudes are damped and nature frequency are enhanced with sufficient lubrication. Vibration and noise of bearings increase obviously and roar can happen and the contact surfaces scratch slightly under insufficient and unclean lubrication state. Bearing will be disabled in a few minutes without any lubrication.4. Bearing vibration is unstable under low speed. With the speed increasing, the vibration become stable and natural frequencies increase slightly but the amplitudes increase apparently.5. Bearing vibration can be excited only by some harmonics ofwaviness without lubrication and by all harmonics with lubrication based on the theoretical analysis.6. The nature frequencies of ball bearings decrease with the increasing sizes. When international clearance increasing, nature frequencies decrease in radial and increase in axial and angular and the amplitudes increase in radial and decrease in axial and angular. When the number of balls increasing, nature frequencies increase and amplitudes decrease. When the pitch diameter increasing, axial nature frequency increase and others nearly keep no change. When outer race groove curvature radius increasing, nature frequencies increase in radial and decrease in axial and angular and amplitudes keep no change in radial and increase in axial and angular. When inner race groove curvature radius increasing, nature frequencies decrease and amplitudes increase.7. The distortion in amplitude and frequency components of bearing vibration signals picked up by the present probe measurement method is founded and right conclusions cannot be achieved by the signals. When accelerometer is rigidly screwed with steel stud onto a flat outer surface of a ball bearing, vibration signal can be got without distortion and the reliability of research on bearing vibration is assured.8. AR model is suitable for large samples of bearing vibration signals, square root of length of samples can be used as the upper limitation of order determination and the FPE order selection criterion is effective. Many advantages of AR spectrum are founded over the classical based on FFT.9. Time-Frequency analysis is necessary for abnormal noise of ball bearings. WPT overcomes the principle shortcomings of STFT and proved to be a best tool to process the abnormal signals.

理论分析和试验研究表明： 1球轴承振动是钢球—滚道接触副中的各种摩擦学作用引起的，具有摩擦动力学的本质，任何对接触副的摩擦学特性有影响的因素都将对球轴承的振动和噪声特性产生影响； 2中心轴向载荷作用下，载荷增大使球轴承的固有振动频率升高，载荷越大同样的载荷增幅引起的频率升幅减小，足够大的中心轴向载荷作用下载荷的变化对球轴承振动的频率特性不会产生明显的影响； 3轴向和径向联合载荷作用下，径向载荷不大时球轴承振动的固有频率基本不变，但是可能引起变刚度激励的非线性振动，出现新的频峰，过大的径向载荷将使部分钢球脱离接触，使球轴承的振动和噪声呈现不稳定状态； 4润滑对球轴承的振动和噪声特性有重要的影响，良好的润滑对振动有明显的抑制作用，润滑不充分时，振动和噪声的水平会有明显增高，一定条件下还会激发接触副中的谐振，发出啸声，造成接触表面的伤害，无润滑干接触时，短时的运转就会损伤接触副表面，使振动和噪声迅速增大，并随时可能引发严重的磨损和卡滞失效；接触副润滑良好时，油膜的"刚化效应"使球轴承振动的固有频率有所提高，润滑剂中含有弥散性污浊时，振动的幅度总体升高，但弥散性污浊不会改变球轴承振动的频率特性；浙江大学博士学位论文：球轴承振动的研究 5低速时，球轴承振动的基本特征呈现不稳定状态，随着转速的提高，球轴承振动的频率特性趋于稳定，固有频率频峰升高； 6理论分析表明，干接触时钢球和滚道表面波纹度的某些谐波分量能激励球轴承的振动，振动的幅值与谐波幅值成正比，实际球轴承中钢球的分布不可能完全均匀，波纹度的激励作用会随时发生；润滑接触的分析表明，波纹度的任意谐波分量均能激励球轴承的振动； 7球轴承的几何和结构参数分析表明，球轴承尺寸越大，径向、轴向和角振动的固有频率越低；钢球中心圆直径增大，球轴承的径向和轴向振动的固有频率基本不变，角振动的固有频率有所上升，振动的幅频特性基本不变；径向游隙加大使球轴承的径向振动固有频率降低，轴向和角振动的固有频率有所升高，径向振动幅频特性的幅值升高，而轴向和角向降低；钢球的数量增多使球轴承振动的固有频率上升，幅频特性的幅值下降，径向振动的幅值下降最为明显；外圈沟曲率半径系数增大使球轴承径向振动的固有频率升高，轴向和角振动的固有频率降低，径向幅频特性基本不变，轴向和角向幅频特性幅值升高；内圈沟曲率半径系数增大使径向、轴向和角振动固有频率均下降，振动幅频特性的幅值均有升高； 8试验对比表明，传感器采用探针式安装时，由于探针接触副接触特性的影响，钡(量得到的球轴承振动信号有失真，采用专门设计和制作的试验球轴承，以固定式安装加速度计，首次测量得到了球轴承振动的真实信号，通过对振动信号的分析，验证了球轴承振动的理论模型； 9)基于时间序列分析的AR模型适用于大样本球轴承振动信号的分析，以样本长度的均方根值作定阶上界，FPE做判阶准则，给出的AR谱光滑，频率分辨率高，是球轴承振动分析的简便而可靠的手段； 10以时频域分析的小波包变换分析球轴承异音信号能够比较好地定位和聚焦异音发生的时间，时间间隔，频率范围，同样是球轴承振动分析的可靠的手段，可用作球轴承故障诊断技术。

Regarded taxi as the source of floating car data,aimed at the disequilibrium of distribution of FCD in road network,two indexes named detecting intensity and detecting rate were designed to analyze the distribution characteristics of FCD in different areas and traffic corridors by large sample analysis.

以出租车作为浮动车数据采集源，针对浮动车数据路网分布的不均衡性，提出覆盖强度和覆盖率2个指标，通过大量样本分析浮动车数据路网分布的时变特征，以及在不同区域、不同通道上的分布特征。

We deduce the precise distribution of Pearson χ〓 in small sample case and our calculation is compared with the monte-carlo simulation results.

在统计样本很小时，常用的皮尔逊χ〓量的分布将会偏离χ〓分布，这时如果仍采用χ〓检验将会极大地影响检验结果，我们给出了这种情况下的皮尔逊χ〓量的准确分布并与蒙特卡洛模拟结果作了比较。

In the empirical analysis part, statistical descriptions on the sample funds are made firstly, which show that the actual distribution of each sample fund"s daily net value return possesses the characteristic of leptokurtosis. So it is necessary to add student T distribution and GED to capture such leptokurtosis characteristic other than normal distribution. Secondly, ARCH test shows that there exists volatility clustering in each sample fund"s daily net value return, so GARCH related models should be used to describe such volatility clustering characteristic.

实证分析部分首先对样本基金进行统计描述，得出其收益序列均存在尖峰厚尾特征，不服从正态分布，因此有必要在下面的VaR计算中加入T分布和GED分布来捕捉这种尖峰厚尾特征；并且经ARCH检验后得出收益序列存在明显的波动聚集性的特征，因此可以选择GARCH类模型来描述这种特性，经过模型筛选，得出最适合我国开放式股票型基金的收益波动性模型为GARCH(1，1)模型。

The papes utilizer Parzen's kernel estimate theory which was used in estimating the function ofprobability intensity and Chen Xiru's results in the field to strictly prove the following fact.Whenusing non-completeness knowledge sample W to find the law which W would be kept to,thefrom:is closer to real pthan the estimate of rectangular diagram.

本文借鉴Parzen提出的概率密度函数的核估计理论和陈希孺教授等研究的结论，用概率论的方法严格证明了下列事实：用非完备知识样本估计其观测值应遵循的概率分布规律时，扩散型估计：得到的分布比用直方图法得到的估计更接近于真实概率分布p。

Throught adaptive algorithm of network, the system constantly revise right values distribute so as to meet goal. Distribute the heuristic knowledge and the experience knowledge to network interconnection and right value.?The reasoning machine carries on reasoning according to the knowledge that involve in knowledge base and fact that users offer.?Through front computation of nerve network, the output port obtains the output vector.?The explanation module of ES is explained to user.At last, this paper has realized grab hydraulic system based on ANNES intelligent diagnosis, it has proven this fault diagnosis method feasibility.

ANNES的工作过程是：①首先建立知识库；②学习系统获取知识时，它只要求专家提供范例及相应的解，通过神经网络学习算法对样本进行学习，经过网络内部自适应算法不断修改权值分布以达到要求，把专家求解实际问题的启发式知识和经验分布到网络的互连及权值分布上；③推理机根据知识库中的知识和用户提供的事实进行推理；④通过神经网络的前向计算，在输出端获得输出矢量：⑤经专家系

However, as the name（read－only memory）implies, CD disks cannot be written onorchanged in any way.

然而，正如其名字所指出的那样，CD盘不能写，也不能用任何方式改变其内容。

Galvanizes steel pallet is mainly export which suits standard packing of European Union, the North America. galvanizes steel pallet is suitable to heavy rack. Pallet surface can design plate type, corrugated and the gap form, satisfies the different requirements.

镀锌钢托盘多用于出口，替代木托盘，免薰蒸，符合欧盟、北美各国对出口货物包装材料的法令要求；喷涂钢托盘适用于重载上货架之用，托盘表面根据需要制作成平板状、波纹状及间隔形式，满足不同的使用要求。