新疆天山地区 PP回归综合预报模型研究及预报效能评价

运用基于 PP回归理论的数值型地震综合预报软件系统, 对新疆天山地区进行了 PP回归建模研究。 通过回顾性 PP回归动态建模预测检验,对其预报效能进行了评价。 结果认为,该模型作为时序性数值预报模型具有较好的中短期预报效能。 尤其对地震样本量大、震级分布完备的地区,其预报效能更好。     

人工神经网络在地震综合预报中的应用

多级判别、聚类分析等统计模式识别方法在地震综合预报中的应用已有多篇文献报道(石绍先、沈斌,1986;1990;王学仁,1982).多年预报实践表明,模式识别方法对于地震预报是有效的.随着国际上神经网络研究的兴起,该方法也被用于模式识别领域,但尚未见其用于地震综合预报方面.本研究运用“反向传播”神经网络,对滇西南地震区17个震例进行了计算机分析,建立了地震综合预报的专家系统,从而为进一步研究决定地震强度的物理因素,准确地预报将来的未知地震提供参考信息.     

BMA集合预报在淮河流域应用及参数规律初探

以淮河流域吴家渡水文站作为试验站点,采用基于贝叶斯平均法(BMA)的集合预报模型处理来源于马斯京根法、一维水动力学方法、BPNN(Back Propagation Neural Network)的预报流量序列,通过分析BMA的参数以及其预报结果,对各方法在淮河典型站点流量预报中的适用性进行验证与分析.经2003—2016年19场洪水模拟检验可知,BMA模型能够有效避免模型选择带来的洪水预报误差放大效应,可以提供高精度、鲁棒性强的洪水预报结果.通过进一步比较各模型统计最优的频率与BMA权重值之间的相关性,发现权重值不适用于对单场洪水预报精度评定,而适用于描述多场洪水预报中,模型为最优的统计频率;基于大量先验信息,提前获取BMA的权重等参数,将是指导模型选择、降低洪水预报不确定性、改进洪水预报技术的有效手段.     

预报手段效率值的研究

对一种预报手段的科学评价,首要的是区分它是否是一种有效非盲目的预报手段另一个重要方面就是对有效预报手段科学地给定标志预报水平,或有效程度的效率值,以便比较各种不同预报手段的优劣.前者我们在朱令人等(1990)和朱成熹等(1989)利用统计检验的思想已得到解决.本文利用统计检验中的统计量,构造了 Z 系列效率值,通过对现有各种效率值的分析比较,证明了它具有较多的优点,是目前最好的评分法之一.      

神经网络在地震学方法综合预报中的应用

将BP神经网络用于地震中期预报．使用一些常用的地震学指标作为神经网络的输入,而将BP神经网络的输出作为表征地震活动增强的特征参数W0,并将其用于华北地区进行空间扫描．结果表明,中强地震前1～3年,未来震中周围通常开始都出现明显的W0值中期异常区．本方法具有很好的中期预报效果．      

非平稳时间序列的动态水位神经网络预报模型

水文预报系统是一个复杂的非线性动力学过程,站点水位受各种因素的影响不仅呈现出非平稳动态随机变化特性,而且各因素间的关系也很难确定。淮河流域五河站水位由于受到洪泽湖回水影响及季节性的影响,也呈现出这一动力学的非平稳特性,因此本文在考虑了相关站点和回水影响的基础上,建立了一种多站变量时间序列的神经网络预报模型,预报结果表明该方法预测效果较好,运行简单。     

月平滑太阳黑子相对数的预报

建国以来,不少天文工作者在太阳活动长期预报、中期和短期预报方面作了大量工作,但是,太阳活动的一至几年的中长期预报却很少有人涉足。本文试图用门限自回归模型和二次曲线方法,探讨时间长度为一年的月平滑黑子相对数的预报。并给出两种方法的综合预报结果。     

神经网络在短期地震跟踪预报中的应用

本文研究了神经网络方法在基于地震活动性指标的中短期地震预报和基于非测震学前兆异常从属函数的短期地震预报中的应用。选用含一个或两个中间层的前向神经网络模型，并采用与之相适应的ＢＰ算法。以华北地区多年的地震活动性资料和首都圈及其邻近地区的短水准、地电阻率、地磁总强度、水位、水氡含量等前兆观测手段的８０余个台项的多年测资料为基础，对神经网络方法以上两方面的应用作出了实际计算、分析与检验。对一些大地震的发     

新疆主要地震区pp回归综合预报模型研究

应用基于ｐｐ回归理论的数值型地震综合预报软件系统,选择了ｂ值、Ｃｂ值、Ａｂ值等１５个测震学参量,建立了它们与响应变量（未来任意时段内最大地震震级）的ｐｐ回归模型．对新疆４个主要地震区分别给出了各区的建模岭函数及综合预报模型．回顾检验及实际预报结果均表明该模型的预报效果是较为理想的．     

灰色系统预测在地电预报地震三要素中的应用

本文用灰色系统理论中的系统预测方法，分别建立震级Ｍ、震中距Ｒ、发震时间ｔ＿ｆ与地电异常幅度Ｕ、持续时间Ｔ的灰色预测模型ＣＭ（１，３）。由于引入了ＧＭ模型，就避免了人为确定统计函数形式，并且在震例较少的情况下也可以进行三要素预报，这样就可以建立起适合各个具体台站的预测关系，预测结果较目前统一的预报公式要好。该方法所需资料少，计算简单，给出的结果简明实用。     

Improved ENSO simulation in regional coupled GCM using regressive correction method

A regressive correction method is presented with the primary goal of improving ENSO simulation in regional coupled GCM. It focuses on the correction of ocean-atmosphere exchanged fluxes. On the basis of numerical experiments and analysis, the method can be described as follows: first, driving the ocean model with heat and momentum flux computed from a long-term observation data set; the pro-duced SST is then applied to force the AGCM as its boundary condition; after that the AGCM’s simula-tion and the corresponding observation can be correlated by a linear regressive formula. Thus the re-gressive correction coefficients for the simulation with spatial and temporal variation could be obtained by linear fitting. Finally the coefficients are applied to redressing the variables used for the calculation of the exchanged air-sea flux in the coupled model when it starts integration. This method together with the anomaly coupling method is tested in a regional coupled model, which is composed of a global grid-point atmospheric general circulation model and a high-resolution tropical Pacific Ocean model. The comparison of the results shows that it is superior to the anomaly coupling both in reducing the coupled model ‘climate drift’ and in improving the ENSO simulation in the tropical Pacific Ocean.     

地震预报中地电阻率高精度计算方法

本文利用回归分析、谐波分析和预测、理论研究了从大地电位差(Vsp)和人工电位差(ΔV)迭加的合成势差(E)中高精度分离ΔV的方法。实际计算表明:当Vsp变化规律较强时,该方法的精度比常规方法提高几倍至十几倍;当Vsp变化规律较弱时,该方法的精度一般高于或接近于常规方法的精度。     

地震波时域数值优化研究及应用

将最小二乘拟合法与单自由度地震动力反应递归法相结合来解决地震观测和人工模拟地震波过程中速度、位移的基线漂移问题。首先以3次多项式来拟合加速度的均值线(一次优化);再对一次优化后的积分速度、位移时程仍存在的长周期基线波动问题,运用单自由度地震动力反应递归法进行二次优化。文中通过3个数值算例体现出优化算法的优越性,并将优化后的地震波加速度时程应用到江坪河水电站溢洪道控制段三维有限元动力计算分析中。结果表明,该算法消除了积分基线漂移影响,具有较好的可行性、数值稳定性和易操作性。     

回归分析方法在地震科学应用中的问题与探讨

分析了回归分析方法在地震科学应用中存在的问题,提出了按变量之间的物理关系进行分类回归的新观点。针对实际地震问题给出了具体的分类方案,从理论上论述了不同类型问题采用不同的回归方法的必要性。通过对实际地震问题分析,剖析了现有回归分析方法的优缺点和适用条件。在此基础上,对不同类型统计问题如何选用适当的回归分析方法提出了合理的建议。     

Active sensing using impedance‐based ARX models and extreme value statistics for damage detection

In this paper, the applicability of an auto‐regressive model with exogenous inputs (ARX) in the frequency domain to structural health monitoring (SHM) is established. Damage sensitive features that explicitly consider non‐linear system input/output relationships are extracted from the ARX model. Furthermore, because of the non‐Gaussian nature of the extracted features, Extreme Value Statistics (EVS) is employed to develop a robust damage classifier. EVS provides superior performance to standard statistical methods because the data of interest are in the tails (extremes) of the damage sensitive feature distribution. The suitability of the ARX model, combined with EVS, to non‐linear damage detection is demonstrated using vibration data obtained from a laboratory experiment of a three‐story building model. It is found that the vibration‐based method, while able to discern when damage is present in the structure, is unable to localize the damage to a particular joint. An impedance‐based active sensing method using piezoelectric (PZT) material as both an actuator and a sensor is then investigated as an alternative solution to the problem of damage localization. Copyright © 2005 John Wiley & Sons, Ltd.     

A simple overland flow calculation method for distributed groundwater recharge models

A method to improve the calculation of overland flow in distributed groundwater recharge models is presented and applied to two sub‐catchments in the Thames Basin, UK. Recharge calculation studies tend to simulate the runoff flow component of river flow in a simplistic way, often as a fraction of rainfall over a particular period. The method outlined in this study intends to improve the calculation of groundwater recharge estimates in distributed recharge models but does not present an alternative to complex overland flow simulators. This method uses seasonally varying coefficients to calculate runoff for specified hydrogeological classes or runoff zones, which are used to model baseflow index variations across the basin. It employs a transfer function model to represent catchment storage. Monte Carlo simulation was applied to refine the runoff values. Decoupling the runoff zones between the two sub‐catchments produces a better match between the simulated and observed values; however, the difference between observed runoff and the simulated output indicates other factors, such as landuse and topographical characteristics that affect the generation of runoff flow, need to be taken into account when classifying runoff zones. British Geological Survey © NERC 2011. Hydrological Processes © 2011 John Wiley & Sons, Ltd.     

地震作用下结构时变物理参数识别

利用小波分析函数多尺度逼近方法,将剪切型结构在地震作用下时变的阻尼和刚度用尺度函数的线性组合表示,将时变参数的辨识问题转化为由已知的正交尺度函数和系统的输入输出来估计线性组合中的时不变系数问题,用最小二乘法对剪切型框架结构在地震作用下时变的刚度和阻尼进行了有效的辨识。此方法无需事先假定系统参数的时变规律,在有噪声情况下可以用Tikhonov正则化方法减小识别方程的不适定性对识别结果的影响。数值算例表明了该方法的有效性。     

地震波数值模拟的非规则网格PML吸收边界

以格子法为基础,以声波方程为例研究非规则网格PML(Perfectly Matched Layer)方法.本方法的核心是建立局部坐标系下的分裂方程和基于积分近似的微分方程弱形式.该非规则网格模拟方法允许在计算域内设置任意形状的人工边界.对于二维半空间问题,与采用矩形人工边界相比,采用半圆形人工边界可减少计算量20%以上.采用光滑的曲边界,不仅可减少计算区域,还可避免常规的PML吸收边界在吸收带角点区域的特殊处理.本方法事先计算和存储边界单元的局部几何参数,在计算的每一时间步查表调用这些参数,与常规的直边界PML方法相比,不增加任何计算量.     

Algorithms for time synchronization of wireless structural monitoring sensors

Dense networks of wireless structural health monitoring systems can effectively remove the disadvantages associated with current wire‐based sparse sensing systems. However, recorded data sets may have relative time‐delays due to interference in radio transmission or inherent internal sensor clock errors. For structural system identification and damage detection purposes, sensor data require that they are time synchronized. The need for time synchronization of sensor data is illustrated through a series of tests on asynchronous data sets. Results from the identification of structural modal parameters show that frequencies and damping ratios are not influenced by the asynchronous data; however, the error in identifying structural mode shapes can be significant. The results from these tests are summarized in Appendix A. The objective of this paper is to present algorithms for measurement data synchronization. Two algorithms are proposed for this purpose. The first algorithm is applicable when the input signal to a structure can be measured. The time‐delay between an output measurement and the input is identified based on an ARX (auto‐regressive model with exogenous input) model for the input–output pair recordings. The second algorithm can be used for a structure subject to ambient excitation, where the excitation cannot be measured. An ARMAV (auto‐regressive moving average vector) model is constructed from two output signals and the time‐delay between them is evaluated. The proposed algorithms are verified with simulation data and recorded seismic response data from multi‐story buildings. The influence of noise on the time‐delay estimates is also assessed. Copyright © 2004 John Wiley & Sons, Ltd.     

独塔刚构体系斜拉桥地震反应分析

基于非线性有限元理论,以某独塔双索面刚构体系斜拉桥为例建立了动力空间有限元模型,对该桥的动力特性进行了研究。在此基础上通过三角级数拟合规范反应谱的方法,合成了适用于结构分析的人造地震波,并以此和两条实际地震记录作为输入地震动,应用时程分析法对比分析了一致激励和行波激励下的结构地震反应。研究结果表明,在考虑纵向+竖向组合作用时,独塔双索面斜拉桥的内力和位移反应比纵向分量单独作用时更为显著;行波效应可以明显减小结构的位移反应,对结构抗震来讲是有利的。以上结论将为此类桥梁的设计和发展提供一定的理论依据。