地震层析成像的实用正则化

正则化是求解地震层析成像反演问题必须的。一般情况下,地震层析成像方程组非常庞大,合理选择正则化参数非常困难。本文提出一种在线性层析成像反演中实际选择正则化参数的算法。该算法基于大多数层析成像常用的数据统计假设。首先利用Lanczos双对角变换将方程组转换到Krylov子空间。在转换的子空间中,方程组转换成标准阻尼最小二乘正规方程。标准方程组的解可以写成正则化参数的显函数,这使得可以进行计算机方便的选取正则化参数。正则化参数选取的第二个准则是数字化模拟的研究。如果转换空间的维数远小于原始模型空间的维数,此算法的计算效果非常好,特别适用于大型地震层析成像问题。     

多层体柱微波成像的连续编码遗传算法

提出一种能同时重建多个二维良导体目标外形轮廓的新方法－－连续编码遗传算法,将目标的横截面轮廓近似表达为三角级数形式,由边界条件得到一积分方程组,在此基础上将成像问题转化为约束优化问题,级数的各项系数为待优化量,积分方程组为约束条件,目标函数定义为实际测量的散射场与反演过程中得到的散射场之间的相对误差函数,采用连续码遗传算法求解,优化过程通过选择,交叉、变异等遗传操作的迭代而实现,待优化目标函数进行     

多导体柱微波成像的连续编码遗传算法

提出一种能同时重建多个二维良导体目标外形轮廓的新方法——连续编码遗传算法．将目标的横截面轮廓近似表达为三角级数形式,由边界条件得到一积分方程组,在此基础上将成像问题转化为约束优化问题,级数的各项系数为待优化量．积分方程组为约束条件,目标函数定义为实际测量的散射场与反演过程中得到的散射场之间的相对误差函数．采用连续编码遗传算法求解,优化过程通过选择、交叉、变异等遗传操作的选代而实现,待优化目标函数进行线性变换并采用模拟退火原理确定目标后表达为适应度函数,采用联赛选择与比例选择相结合的选择机制和单点交叉方式,变异操作通过对基因施加微小随机扰动实现．上一代中适应度最高的个体直接保留．数值模拟反演实验验证了方法的有效性．与其他反演方法相比,本法具有简单、通用、鲁棒性强等特点．     

基于Shearlet变换和广义全变分正则化的地震数据重建

压缩感知技术通常利用地震信号在某一变换域内的稀疏性质,将随机缺失的地震数据重建问题转化为L1正则化问题.本文首先通过Shearlet变换获得地震信号的稀疏性质,再将广义全变分（TGV）约束引入L1正则化模型,构建了基于Shearlet变换的双正则化模型用于重建地下介质的图像.与传统L1正则化方法相比,基于Shearlet变换的双正则化方法不仅考虑了信号的稀疏性,同时兼顾了地下介质结构的复杂性,可以较好的重建地下结构体的图像.最后采用交替方向乘子法（ADMM）求解所建模型,每个子问题均可得到显式解.数值实验对比了基于小波变换、Shearlet变换的L1正则化方法和TGV正则化方法,结果表明基于Shearlet变换的双正则化方法对于随机采样50%数据的情况具有较好的重建结果,同时对于有限范围的连续缺失数据的重建亦具有一定的有效性.     

基于Shearlet变换和广义全变分正则化的地震数据重建

压缩感知技术通常利用地震信号在某一变换域内的稀疏性质,将随机缺失的地震数据重建问题转化为L1正则化问题.本文首先通过Shearlet变换获得地震信号的稀疏性质,再将广义全变分(TGV)约束引入L1正则化模型,构建了基于Shearlet变换的双正则化模型用于重建地下介质的图像.与传统L1正则化方法相比,基于Shearlet变换的双正则化方法不仅考虑了信号的稀疏性,同时兼顾了地下介质结构的复杂性,可以较好的重建地下结构体的图像.最后采用交替方向乘子法(ADMM)求解所建模型,每个子问题均可得到显式解.数值实验对比了基于小波变换、Shearlet变换的L1正则化方法和TGV正则化方法,结果表明基于Shearlet变换的双正则化方法对于随机采样50%数据的情况具有较好的重建结果,同时对于有限范围的连续缺失数据的重建亦具有一定的有效性.     

求解非线性阻尼作用地震破裂过程的一种方法

把所有的参数都无量纲化,然后在证明了无量纲的非线性项的阻尼(与运动速度立方正比)系数B<1为小参数之后,把待求解的振幅函数展成B的收敛幂级数,再代入动力方程或方程组,得到一系列线性化的逼近方程或方程组,用以动坐标表示的广义富氏级数代入各级逼近方程或方程组,对Ⅰ、Ⅱ、Ⅲ型破裂的弹性断裂动力学问题和Voigt模型的粘弹性断裂动力学问题,都化为含时间自变量的Mathieu方程或方程组,用WKBJ方法得出其在稳定区域的渐近解。还用此法解了一个三维问题。     

地震数据反假频规则化方法研究

由于诸多因素的影响,地震数据沿空间方向通常是稀疏采样的,因此引起较为严重的空间假频.本文提出一种反假频地震数据规则化的方法,采用Fourier变换域加权范数带限重建方法完成低频数据重建,利用自适应频谱加权范数的正则化项约束方程的解,将地震数据的带宽和谱形状作为先验信息,具有较好的低频重建特性.文中采用共轭梯度算法求解方程,而后利用重建的低频数据信息,应用频带延拓的方法重建高频数据,未知的高频带信息由重建的低频带信息构建.本方法在完成地震数据规则化的同时,可有效去除地震数据中的空间假频干扰.理论模型和实际资料处理均表明文中所提出的反假频地震数据规则化方法是有效可行的.     

大尺度大气方程组解和吸引子的存在性

首先采用了适当的函数空间, 引入了合理的算子表达形式, 将复杂的大尺度大气动态方程组用一个简单的抽象的算子方程表示, 由此给出了大气方程组弱解的定义. 然后利用Galerkin方法证明了大气动态方程组弱解的存在性和平衡态方程组弱解的存在性, 通过构造大气动态方程组轨道吸引集, 最终得到大气方程组按Chepyzhov-Vishik定义下的非空的整体吸引子的存在 性, 为进一步讨论大气吸引子的拓扑结构和动力学行为建立了基础. 采用的方法对其他大气动态模型解的存在和全局渐近行为的研究也是适用的.     

基于Abel变换的图像重建自适应方法

本文论述了利用轴对称物体的单幅投影信息进行密度重建的一种自适应正则化模型。所提模型基于全变分正则项与高阶全变分正则项的联合使用,主要的优点是在保持清晰的界面及恢复平稳变化区域的同时减弱了阶梯效应。并且使用自适应方法,提高了效果的同时简化了所使用的参数。对于其中涉及的最优化问题,我们采用增广拉格朗日方法来解。数值结果表明,这一模型提高了关于密度界面位置及密度值的准确度,具有较好的抗噪性。      

基于深度卷积神经网络的地震数据重建

地震数据重建是一个不适定的反问题,通常采用正则化方法求解.正则化方法需要人工建模,建模的准确性会影响重建结果,此类方法还存在计算代价高的问题.为克服正则化方法存在的问题,本文使用深度卷积神经网络实现了端到端的地震数据重建.此方法是基于数据驱动的,直接从数据中学习输入与输出的映射关系,无需人工建模,经过训练的网络可直接用于非完整数据的重建工作.数值实验分别使用模拟数据和实际数据并与传统正则化方法对比验证深度卷积神经网络方法的有效性.实验结果表明,深度卷积神经网络方法的计算代价主要在于网络的训练阶段,数据重建阶段仅需花费极短的时间,与传统正则化方法相比,对于缺道50%的地震数据,深度卷积神经网络方法的重建结果质量更高,速度更快.     

Optimisation of the Operation of an Electrochemical Process To Treat TBT‐contaminated Sediments on a Pilot Scale

Since at least three decades tributyltin had been commonly used in paint formulations to protect ship hulls and submersed surfaces from fouling. This has resulted in negative and Technology, effects to the marine environment due to its high toxicity towards non‐target organisms and to its accumulation in sediments, especially near point sources of pollution. Even if a ban of TBT for its application in anti‐fouling paints is enforced the problem of TBT‐contaminated sediments will remain for the coming years due to the low degradation rates of organotin in sediments. In this work an electrochemical process at a pilot scale to destroy TBT in sediments was investigated. Previous experiences at a laboratory and technical scale have demonstrated the potential of this process to electrolytically degrade organotin species in sediments. This article deals with the optimisation of the pilot plant performance in terms of organotin degradation and operation costs monitoring additionally the effect of the electrochemical treatment on other pollutants such as heavy metals, PCBs, and PAHs. The aim was to investigate by means of a parametric study the range of operating conditions required to destroy TBT down to concentrations lower than 100 μg/kg; to determine the influence of those parameters on the performance of the process; and to evaluate the operation costs associated to the process. It was determined that it was possible to destroy TBT down to 100 μg/kg at current densities in the range of 4.4 to 6.6 mA/cm² and chloride concentrations of least of 0.2 mol/L with operation costs of 13 to 18 EUR/m³.     

The effects of salinity on the Manila clam (Ruditapes philippinarum) using the neutral red retention assay with adapted physiological saline solutions

This study investigated the internal osmotic regulatory capabilities of the Manila clam (Ruditapes philippinarum) following in vivo exposure to a range of salinities. A second objective was to measure the health status of the Manila clam following exposure to different salinities using the neutral red retention (NRR) assay, and to compare results using a range of physiological saline solutions (PSS). On exposure to seawater of differing salinities, the Manila clam followed a pattern of an osmoconformer, although they seemed to partially regulate their circulatory haemolytic fluids to be hyperosmotic to the surrounding aqueous environment. Significant differences were found when different PSS were used, emphasizing the importance of using a suitable PSS to reduce additional osmotic stress. Using PSS in the NRR assay that do not exert additional damage to lysosomal membrane integrity will help to more accurately quantify the effects of exposure to pollutants on the organism(s) under investigation.     

A Basic Water Quality Model for the River Neckar: Part 2 – model‐based analysis of the oxygen budget and scenarios

In this paper the Basic Water Quality Model (BWQM) for the central part of River Neckar is used to analyse the oxygen budget and to assess the potentials of various measures to prevent or mitigate critical dissolved oxygen (DO) declines. It is shown that the oxygen budget is mainly governed by phytoplankton dynamics. The excessive growth of algae and the sudden break down of the resulting algal blooms may cause episodic DO depressions. Therefore, to stabilise the oxygen budget in a sustainable way, eutrophication has to be controlled within the central part of River Neckar and the upstream regions. The only feasible way to reach this goal appears to be a further drastic reduction of phosphorus emissions. In addition, it is indispensable to hold the very high standards of biochemical oxygen demand and ammonium retention at the wastewater treatment plants. A worse performance of the treatment plants would dramatically aggravate critical DO declines which may be caused by algae dynamics. As long as the oxygen budget is not completely stabilised, weir and turbine aeration can be used to mitigate DO depressions. It could be shown that the potentials of these measures suffice to keep DO at a tolerable level. However, due to the long travel times in River Neckar, it is important to start aeration up to several days before the DO minimum is reached.     

Prediction of the location of future rupture surfaces of a slowly moving loess landslide by electrical resistivity tomography

A slowly moving loess landslide along the River Danube in South Hungary was studied using electrical resistivity tomography. Our aim was to determine the fracture system of the study site. Due to the homogeneous composition of the loess, it seems to be the only possibility to get information about the landslide and its further evolution. The applicability of the electrical resistivity tomography technique for such a supposedly dense fracture system was studied by numerical modelling, and the results have been verified in the field. The dip of the fractures could not always been observed, and they could not be explored deeply. However, it was possible to map their surface projection to get the desired information about the structure of the landslide. Fracture zones could have been especially well localized, enabling the prediction of the positions of future rupture surfaces and thus the delineation of the endangered zone. Although the area outside of the already subsided one is not endangered yet, the area which has already started to move is going to break into two. Parts of the about 5 m wide blocks at the front of the landslide may fall or slide down anytime. A large area was assumed to move as one unit. Most of our predictions have been verified by the mass movements that occurred about one and half years after the measurements. The electrical resistivity tomography method proved to be a good tool to characterize the fracture system of such a landslide area, enabling the prediction of future rupture surfaces and also delineation of the endangered area. Its use is therefore highly recommended to monitor landslides to provide early risk warnings to avoid damage to constructions or endangering human life.     

地震空区的定量判据

在中国，孕震空间已经比较广泛地应用于地震预报的实践，并多次取得实际预测强震的效果。但经过系统震例研究发现，利用孕震空间预报地震的虚报漏报率比较高。本文运用解决临界相变问题的重正化群方法所求出的孕震空间区可能发生失稳破裂，导致强震发生的判据对孕震空间的危险性作出进一步鉴别，并用突变论方法求解尖拐空变方程，预测孕震空区未来发生强震的时间和震级，并进一步确定孕震空区的危险性。对２６个中国大陆强有的孕震空     

Improving stability of regional numerical ocean models

An operational limited-area ocean modelling system was developed to supply forecasts of ocean state out to 3 days. This system is designed to allow non-specialist users to locate the model domain anywhere within the Australasian region with minimum user input. The model is required to produce a stable simulation every time it is invoked. This paper outlines the methodology used to ensure the model remains stable over the wide range of circumstances it might encounter. Central to the model configuration is an alternative approach to implementing open boundary conditions in a one-way nesting environment. Approximately 170 simulations were performed on limited areas in the Australasian region to assess the model stability; of these, 130 ran successfully with a static model parameterisation allowing a statistical estimate of the model’s approach toward instability to be determined. Based on this, when the model was deemed to be approaching instability a strategy of adaptive intervention in the form of constraint on velocity and elevation was invoked to maintain stability.     

用于速度自动拾取的路径积分优化法与光顺处理技术

将用于图像处理的光顺技术引用到速度谱能量团曲面的处理环节,借以提高速度谱的分辨率,为最终获得高精确解提供一全新的途径.具体思路是基于路径积分优化法--Viterbi算法所具有的自动搜寻及获取最优解的功能,将其应用于地震资料处理中的速度自动拾取,使其向前做最大"能量团"的积分向后递归计算最优解--叠加速度,但这种解的准确性往往与所定义的目标函数--速度谱的分辨率有关.演算结果表明,Viterbi算法与关顺处理技术的有机结合,不仅提高了速度提取的精度,而且实现了速度的自动拾取,可提高常规地震资料处理的效率,从而可快捷地为叠前深度偏移提供所需初始速度模型.     

Design of efficient base isolation for hammers and presses

Foundations supporting hammers and presses have to withstand powerful short-period impact loads. When designing these foundations, their vibration amplitudes and the forces transmitted to the supporting piles or soil medium must be reduced to meet serviceability and stability requirements. Mounting systems are often used to achieve this goal. The objectives of this paper are to investigate the efficiency of mounting systems for different foundation configurations and to provide practical guidance for their design. A comprehensive parametric study was conducted and the results were used to establish a set of charts for the design of efficient mounting systems.     

A method for error analysis of sediment yields derived from estimates of lacustrine sediment accumulation

The logistical demands of coring lake sediments tend to preclude the replicate coring necessary to establish error estimates for measured sedimentary parameters. However, if such parameters are to be used to reconstruct sediment yield, and particularly to identify temporal variability of sediment yield, reasonable error estimates are required. In this paper data from a series of alpine lakes in British Columbia are applied to develop a new method for deriving such estimates. Regression surfaces fitted to point values of sediment mass are used to model the physically controlled spatial variability of sedimentation. Deviations from these surfaces are assumed to represent remaining unstructured variance, which constitutes a conservative error estimate. Application of the technique to the alpine lake dataset gives sediment yield estimates with error ranges of ±7–21 per cent. The potential error is minimized where the spatial variability of sedimentation is strongly predictable. The best fits were achieved for elongate lakes of simple basin morphology. The range of the error estimates is sufficiently low to allow detection of variability in Holocene sediment yield to one of the lakes. By using this technique, absolute sediment yields with associated error estimates may be derived. The associated gains in precision justify multicore approaches to lake sediment‐based reconstructions of sediment yield. Copyright © 2000 John Wiley & Sons, Ltd.