基于能量分布特征的地震事件自动识别

研究了地震信号在小波包变换下的特性,依据地震事件识别中“历史事例对比法”的思想,根据不同震源地震信号频率时变特性的不同．提出了基于“能量分布特征”的特征值,同时采用该特征值用神经网络方法对地震事件进行识别分类。该方法不依赖于系统的数学模型,而是直接利用各频率成分能量的变化提取特征值作为神经网络的输入特征向量来进行事件的识别,避免了对地震信号、传播途径准确建模的困难,简便、直观地完成了事件的识别。实验证明,该方法的事件识别率可达到99％以上．是一种有效的地震事件识别方法。     

时频分析技术在地震信号初至估计中的应用

时频分析旨在构造一种时间和频率的联合密度函数,继而揭示信号所含频率分量及其演化特性,是传统Fourier分析的新进展、乃至于革命。本文通过对地震信号噪声特性的分析,研究了基于联合时频谱图三重相关的地震信号初至的估计方法,并给出了实验结果和结论。     

基于自相关函数的地震信号自适应端点检测

在分析信号自相关函数的基础上,引用了自相关相似距离的概念,讨论了自相关相似距离对噪声和噪信号的描述,并据此总结出基于自相关函数的地震信号端点检测方法。给出了算法的具体实现步骤,阐述了检测阈值的确定方法,并对此方法以及时域波形包络和人工查走时表地震信号端点检测方法得到的实验检测结果进行了比较。结果表明,即使在较低的信噪比条件下,采用本算法仍能较准确地检测出地震信号的端点位置。     

基于小波分析和遗传算法的沸腾曲线特征点研究

信号的局部奇异性包含了信号的许多重要信息,小波变换突破了傅里叶分析在时域和频域方面的局部化能力的不足,适合对非平稳信号的处理.基于小波分析的奇异性检测原理,利用小波分析的多分辨分析特性对沸腾曲线的临界热流密度点以及最小膜态沸腾起始点等奇异性特征点进行了检测和分析;同时应用搜索能力极强的遗传算法对检测出的各类特征点数据进行优化处理,拟合出相应的半理论半经验公式.预测结果和检测结果与实验值符合很好.     

基于小波包变换的地震事件分类

研究了地震信号在小波包变换下的特性,依据地震事件识别中“历史事例对比法”的思想,根据不同震源地震信号频率时变特性的不同,提出了基于“能量—事件”的地震事件分类方法。该方法不需要系统的模型结构,而是直接利用各频率成分能量的变化来进行事件识别,避免了对地震信号、传播途径准确建模的困难,简便、直观地完成了事件的识别。采用统计方法对实验样本进行处理,分别拟合出某地震台站记录的某地区天然地震、核爆炸事件的特征值和容差范围,并据此对检验信号集进行了检验,实验证明,该方法的事件识别率可达到95％以上。     

基于高分辨率时频分析的核爆地震信号能量统计初至检测算法

核爆地震波检测是禁核试核查的关键环节之一。常用的短时长时能量均值比、赤池信息准则、傅里叶变换等方法虚警率都比较高,为此,提出一种基于高分辨率时频分析的核爆地震信号能量统计初至检测算法。其基本原理是根据核爆地震事件信号与噪声的频域分布差异,采用窄带分频技术将信号时域波形变换到时频域,在各频带计算局部能量均值与全局能量均值之比并进行二值化处理,当结果大于设定的阈值时,认为此频带该时刻属于地震事件信号的一部分,如果在某一阈值下出现二值化结果相同的情况,只需继续增大阈值进行检测,直到最终只有一个时间点检测到频段的数量最多即可,最大值的跳跃点即认为是初值时刻。利用地下核爆产生的地震信号进行实验研究,结果表明本文方法能够在提高检测准确率的同时解决高虚警问题,在低信噪比信号条件下具有较好的适用性与鲁棒性。     

吸附法测定U3O8粉末比表面积

基于金属氧化物粉末在低温及适宜压力下的单分子吸附平衡，本工作以BET理论为依据，以氮气为吸附质，研究液氮温度下U3O8粉末对N2的吸附行为，测量因吸附引起的压力变化，采用定量气体体积标定单分子吸附表面积，确定工作曲线的斜率和截距．测定U3O8粉末样品的比表面积。方法的精密度优于11％。     

基于核探测器信号波形特征的故障诊断研究

核探测器是一种特殊的随机信号转换器,发生故障时,采用传统的人工方法很难及时有效地对故障进行诊断。本文提出了一种针对闪烁体探测器信号波形特征的在线智能故障检测与分类方法,通过分析闪烁体探测器不同故障时的输出信号变化特征,建立了相应的故障模型。使用小波包算法与支持向量机理论分解并提取特征向量,即可判断故障类型。以ST401闪烁体探测器为例,进行了模拟仿真实验。实验结果表明,基于信号波形特征的数字化方法能快速有效地对闪烁体探测器进行故障自动诊断。     

基于核探测器信号波形特征的故障诊断研究

核探测器是一种特殊的随机信号转换器,发生故障时,采用传统的人工方法很难及时有效地对故障进行诊断。本文提出了一种针对闪烁体探测器信号波形特征的在线智能故障检测与分类方法,通过分析闪烁体探测器不同故障时的输出信号变化特征,建立了相应的故障模型。使用小波包算法与支持向量机理论分解并提取特征向量,即可判断故障类型。以ST401闪烁体探测器为例,进行了模拟仿真实验。实验结果表明,基于信号波形特征的数字化方法能快速有效地对闪烁体探测器进行故障自动诊断。     

基于平稳小波变换的毛刺检测算法

核爆炸地震监测技术研究中,数据质量检测是地震数据自动处理的基本内容,毛刺是影响数据质量的主要问题数据。基于平稳小波变换和非线性能量检测算法,给出一种毛刺自动检测算法。平稳小波变换弥补了正交小波变换存在的不足,可以使尺度分解结果的长度和原始数据保持一致,具备时移不变性。非线性能量检测算法可以增强记录中的高频信号,对平稳小波变换的结果应用非线性能量检测算法,提高了记录中毛刺检测的准确性,非常适合连续地震监测数据自动处理的需要。实验结果表明,给出的这种算法特别有利于记录中小毛刺的检测,从而能够减小信号检测的误检率。     

A study on technique to estimate impact location of loose part using Wigner-Ville distribution

Presented in this paper is a method to estimate impact location of a loose part using the Wigner-Ville distribution. The method uses dispersion characteristics of bending waves propagated in a plate. The power propagation velocity and arrival time difference of bending waves related to the dispersion characteristics can be obtained through the transformation of impact signals using the Wigner-Ville distribution. The distance from the impact location to the signal measuring point can be estimated using the information on the power propagation velocity and the arrival time difference of two bending waves. The experimental results show that the proposed method estimates the impact location with relative percentage error within 10% compared with the actual impact location.     

基于FMmlet变换塔形分解的地震信号滤波

研究了基于FM^mlet变换自适应匹配投影塔形分解原理的地震信号滤波方法。该方法在不提供地震信号背景噪声统计特性的前提下,能有效地去除地震信号中的噪声和干扰,实现地震信号的滤波。分别以实验信号和实际地震信号为例进行了实验,并将实验结果与传统的带通滤波结果进行对比,证明该方法不但可以有效地恢复受干扰污染的或淹没于噪声中的真实信号,且滤波效果优于带通滤波的效果。     

基于腔式探头的束团到达时间测量算法优化

腔式探头具有高信噪比和高灵敏度的特性,TM010模式下在近轴附近的腔体输出信号与束流横向位置无关,信号幅度与电荷量呈正比,信号相位由束团到达时间决定,因此非常适用于高分辨率的束团到达时间测量。上海硬X射线自由电子激光装置（SHINE）对束团到达时间的测量精度要求非常高,期望运用基于谐振腔的束团到达时间监测系统,可实现在100 pC电荷量下分辨率好于25 fs的技术指标。对于高Q的谐振腔信号,单点采样的信噪比随信号的衰减而降低,因此肯定存在一最佳的信号处理窗口。为分析最佳处理窗口与相关参数的定量关系,采用数值仿真和束流实验相结合的办法来研究这个问题。分析发现,存在最佳数据处理时间窗口,该参数只与信号衰减时间常数有关,而与系统采样率、系统信噪比、系统模拟带宽无关。在最佳数据处理时间窗口条件下,束流到达时间测量不确定性可获得最佳值。     

基于连续小波变换的Hilbert包络线松动部件定位方法

本文结合时-频域信号处理方法,在对冲击信号进行连续小波变换(CWT)的基础上,采用计算信号Hilbert包络线极大值的方法确定冲击信号的到达时间,并应用于网格定位。研究结果表明,该方法具有较高的定位精度,对小质量的松动部件能准确地定位其位置,且具有较好的抗噪能力。     

Loose-part signal properties

Lamb's solution for two-dimensional wave propagation was used to identify plate bending wave characteristics observed in loose part impact signals. The surface displacement wave is shown to have large second derivatives at points that can be used to measure the impact contact time. General features of the initial impact wave shape provide guidance for the use of spectral analysis techniques to identify the contact time and wave arrival time when there is a low signal-to-noise ratio. Improvements in the determination of contact time improve the estimation of loose part mass and energy through the Hertz impact parameters. The shape of the initial impact wave shape can also be used to identify on which side of the plate the impact took place. This information has applications for Loose Part Monitoring System metal impact signal validation and for improved estimation of loose part parameters.     

Ballistic deficit compensation method for a large-volume HPGe detector at high count rates

A large-volume HPGe detector normally has a severe ballistic deficit due to its long rise time of the output signals.Despite the trapezoidal shaping algorithm adopted as a remedy to deal with the signals,the algorithm cannot fully eliminate the ballistic deficit in the case of a high counting rate.To resolve the problem,we propose a ballistic deficit compensation method that is based on the measurement of rise time of the signals before shaping.We find that the ballistic deficit after trapezoidal shaping has little relation to the time constant,but shows a quadratic relationship with the rise time and has a negative correlation with the shaping time.In the case of high count rates,the high resolution is handled by fitting the curve to the rise time and the amplitude deficit of the signal after shaping and by compensating for the signal amplitude after trapezoidal shaping.Tests indicate that when the count rate is about 100 kcps,the resolution of Co-60 improves from traditional 2.32 up to 1.91 keV,thus reaching a higher level.     

10MW 高温气冷实验堆的地震监测系统

为了保障地震发生时10MW高温气冷实验堆（HTR－10）的正常运行和安全停堆,设置了地震监测系统。该系统按无人值守方式设计,主要由高灵敏度的加速度传感器、高可靠性的工业控制计算机和成熟的地震分析软件组成,信号采集处理采用了最新技术。系统监测三个测点的九路加速度时程信号,判断为地震信号后计算峰值加速度,超过报警阈值时向主控室提供声光报警信号;系统同时计算响应谱和设计响应谱,并与理论设计响应谱进行比较,最后打印出地震报告。抗震试验和性能试验结果表明,该系统符合抗震设计要求,其功能覆盖并超过了组合使用的多种地震监测仪表。     

Noise and DC balanced outlet temperature signals for monitoring coolant flow in LMFBR fuel elements

Local cooling disturbances in LMFBR fuel elements may have serious safety implications for the whole reactor core. They have to be detected reliably in an early stage of their formation therefore. This can be accomplished in principle by individual monitoring of the coolant flow rate or the coolant outlet temperature of the subassemblies with high precision.

In this paper a method is proposed to increase the sensitivity of outlet temperature signals to cooling disturbances. Using balanced temperature signals provides a means for eliminating the normal variations from the original signals which limit the sensitivity and speed of response to cooling disturbances. It is shown that a balanced signal can be derived easily from the original temperature signal by subtracting an inlet temperature and a neutron detector signal with appropriate time shift.

The method was tested with tape-recorded noise signals of the KNK I reactor at Karlsruhe. The experimental results confirm the theoretical predictions. A significant reduction of the uncertainty of measured outlet temperatures was achieved. This enables very sensitive and fast response monitoring of coolant flow.

Furthermore, it was found that minimizing the variance of the balanced signal offers the possibility for a rough determination of the heat transfer coefficient of the fuel rods during normal reactor operation at power.     

Correction to: GPU-accelerated scanning path optimization in particle cancer therapy

Free electron lasers provide high-power and ultrashort pulses with extreme brightness. In order to improve a facility's capabilities and explore the possibility of performing high-resolution time-resolved experiments, a beam arrival time resolution under 100 fs is required. In this article, a novel beam arrival time monitor(BAM)equipped with two cavities has been designed and a beam flight time measurement scheme based on the BAM prototype has been proposed to estimate phase jitter in the signal measurement system. The two BAM cavities work at different frequencies and the frequency difference is designed to be 35 MHz. Therefore, a self-mixing intermediate frequency signal can be generated using the two cavities. The measured beam flight time shows a temporal deviation of 37 fs(rms).