汶川地震区斜坡动力反应研究
摘要
斜坡是地球表面具有与水平面呈一定倾角的倾斜地质体,可以是天然的,也可以是后期人类活动形成的,它涵盖了滑坡和边坡的概念。滑坡是斜坡破坏型式中分布最广、危害最严重的一种地质灾害。而在所有滑坡灾害中,地震触发的滑坡往往造成更为严重的灾难。汶川大地震灾害调查结果表明,地震诱发的滑坡是主要的地震地质灾害类型之一。研究滑坡动力问题须从斜坡入手,地震动力条件下的斜坡动力反应问题是一个多学科深度交叉的前沿问题：涉及到工程地质、水文地质、构造地质、地震工程、岩土力学等多个学科。
本文以地震条件下的斜坡为研究对象,从动力反应分析这个关键线路进行研究。目的是揭示强震条件下某一类斜坡岩土体稳定性评价问题,服务于生产,造福于人民。
本文采取的总体技术路线是,从一般斜坡静力学问题入手,逐步拓展到斜坡动力分析领域。全文的研究内容如下：
(1)从斜坡稳定性评价方法入手,在前人研究成果的基础上,将斜坡稳定系数的定义分为三大类,即：稳定系数的基本定义,稳定系数的滑动面应力积分定义和稳定系数的强度折减定义,研究三大类稳定系数的内在关系和差别。稳定系数的滑动面应力积分定义与稳定系数的基本定义关系密切,前者是后者在宏观尺度上的自然延伸,定义的内涵是相似的。这两种定义都是在不改变当前应力状态的情况下,按摩擦型材料的观点,以抗剪力与实际剪切应力的比值来定义稳定系数。而强度折减定义,本质上是寻找斜坡的临界破坏状态,以折减系数来描述强度储备,在解决问题的思路上明显不同。在简单块体条件下,强度折减法与基本定义法是等效的,但在复杂的应力场条件下,情况则有很大的不同。为了揭示强度折减法与前两类定义的差别,本文首先探讨应力场与岩土体物理力学性质场之间的联系,继而提出并证明了整体极限平衡状态引理。通过引理从力学上严格证明了基于二分法(含黄金分割法)的强度折减法与滑动面应力积分的定义并不完全等效,真正与之等效的是广义强度折减法。笔者定义的所谓广义强度折减法是指应力场中各点可具有不同的折减系数,由积分中值定理能够证明整体上存在一个等效的强度折减系数,该系数反映的是所有强度折减系数的加权平均,而且该系数与基于滑动面应力积分的定义完全等效。但广义强度折减法的实用价值远不及理论价值,目前尚无有效的实现算法。
(2)研究了强度折减法的适用性,并提出了并行强度折减法。通常基于二分法的强度折减法都是单一折减系数的强度折减,虽然简单易行,但是其等效性不能获得保证。虽然大多数情况,其逼近的特性良好,但不排除在某些情况下很可能会有较大的偏差,特别是在地质条件、几何形态复杂的条件下,各种数值模拟方法捕捉到的临界状态往往是局部的塑性流动,而不是我们所关心的整体极限平衡。因此,在滑动面应力积分定义的框架下开展非圆弧临界滑动面的搜索研究具有非常现实的意义。运用分支预测的思想,提出了二分法的并行多线程算法,并将其应用于强度折减法。详细分析了并行算法的性能,给出了加速比的估算公式。对FLAC3D进行二次开发,将传统的FLAC3D环境下的强度折减法改造成并行的模式,效率提升显著,7个线程条件下的速度提高到原来的近4倍。
(3)从信息学的角度,对滑动面搜索问题进行提炼,严格地证明了斜坡临界滑动面搜索为NPC类问题。证明的意义在于,首先明确了不可能设计出具有多项式时间复杂度的精确求解算法,为滑动面搜索算法的设计指明了方向：正如大量的NPC类问题采用群体智能算法取得很好的效果,采用智能群体算法寻求最优滑动面的近似解是解决问题的大方向。实践证明,采用微粒群算法(PSO)能够获得性能良好的近似解。证明的意义还在于,从信息学的角度印证了基于二分法(含黄金分割法)的强度折减法与常规稳定性系数的定义方法并不完全等效。
(4)研究了基于微粒群算法(PSO)的斜坡非圆弧临界滑动面的搜索方法。以滑动面应力积分的方法来定义斜坡稳定系数,并采用微粒群算法(PSO)来动态搜索临界滑动面。详细阐述了将微粒群算法(PSO)应用于临界滑动面搜索的方法和实现步骤。在此基础上提出了多层次微粒群算法(PSO),为PSO算法与其他优化方法融合奠定了基础。将PSO算法推广到并行处理,大幅度提高了滑动面搜索的效率,实测表明在8线程条件下能实现超过4倍的加速比。此外还详细研究了滑动面的动态随机生成技术。从某种意义上说,将微粒群算法引入一个具体的领域,其难度不在于微粒群算法本身,而在于如何将需要解决的具体问题与算法的框架模型相联系。具体到临界滑动面搜索问题,如何动态地随机生成符合几何约束条件的滑动面是微粒群算法(PSO)能否成功应用的关键。本文以澳大利亚计算机应用协会(ACADS)考核题EX1C为例,详细阐述了滑动面动态随机生成技术、对异常状况的处理方法,以及如何根据具体情况分析,编写高质量的代码,提高效率。
(5)开发了去除仪器响应的接口软件,填补了Windows下缺乏同类软件的空白。该软件能实现自动读取国家地震局公布的SEED/SAC格式文件,运用时域滤波算法去除仪器响应,并以文件形式自动输出速度和加速度时序,可直接供SeismoSignal软件读取进行后续分析。值得注意的是,关于汶川主震波形,目前可供使用的只有速度型地震仪的记录,但速度型地震仪并不能直接记录加速度时程信号,为了从速度时程得到加速度时程,需要采用数值微分算法,所面临的问题是数值微分往往会放大数据中的误差。这就是通常所说的“微分倾向于不稳定”。本文详细阐述和比较了不同的微商算法的精度问题,指出求解实际上等效于延时滤波,阶次越高越明显,与一般问题中倾向于“光滑”的处理恰好相反,加速度的求解应尽可能保留加速度时程的分辨率。同时指出,汶川强震条件下,速度型地震仪饱和现象十分明显,若要获取更为精确的波形则需要采用强震仪进行地震监测。
(6)研究了斜坡动力计算的边界条件的设置、力学阻尼参数的选取方法、网格尺寸和动力荷载滤波的匹配问题。以国家地震局成都台(CD2)记录的波形为动荷载,考察了ACADS考核题EX1C算例在纵、横波双通道耦合输入条件下的动力响应。与常规的岩土动力反应分析中仅仅使用一个通道的波形不同,本文采用纵波和横波耦合输入的方式,能够考虑东西向(EW)、南北向(NS)和垂直方向(H)三个方向上的耦合,尽可能真实的还原研究对象在真实地震条件下的反应。结合斜坡非圆弧临界滑动面搜索的多层次并行微粒群算法,获取了不同时最危险滑动面的位置,以及相应的稳定性系数时程曲线。通过谱分析,指出斜坡稳定系数对动力荷载频率的响应是不均匀的,具有明显的频率选择特性。具体到EX1C算例,对0.8HZ-3.2HZ的范围内的低频动力荷载比较敏感。
(7)开发了FLAC3D和有限元条件下的三维应力场(含物理指标参数场)的快速获取和访问算法。对于复杂的数值模拟问题,速度往往决定一切,否则我们就永远只能停留在解决简单的、概念性的教学模型阶段。与传统的静力分析不同,动力条件下,为了满足频谱分析的采样密度,对应力场获取的时效性提出了更高的要求,以往对此鲜有研究,如何短时间内快速的获取应力场是必须着力解决的问题。为此,笔者以变直接为间接、以空间换时间、改串行为并行的指导思想,分层次、分阶段的逐步提高算法的效率。最终实现了应力场生成和访问的多线程并行算法,获得了显著的效率提升。高速的应力场访问技术为深入进行复杂条件下的大自由度斜坡动力反应分析奠定了基础,使得应力场生成和访问的时效性不再成为制约动力分析的瓶颈,为开展基于应力场的动力反应分析扫清了障碍。有限元条件下与基于有限差分的方法不同,有限元的大部分参数指标保存在网格节点中,情况要更加复杂,本文采用基于体积权重的算法来实现三维条件下应力场的快速获取和访问,并继承了FLAC3D条件下的应力场快速生成技术,实现了多线程并行计算。
(8)把斜坡动力反应分析的基本方法、斜坡非圆弧临界滑动面搜索技术相结合,应用于高速远程滑坡的动力反应研究。以四川省北川县金鼓村滑坡为例,讨论了动力甩出区的时空演化规律,并分析了稳定系数、坡脚和坡顶等各监测点的加速度和速度的频率特性。
本文的主要创新点是：
(1)严格证明了基于二分法(含黄金分割法)的强度折减法与滑动面应力积分的定义并不完全等效,结束了岩土力学界长达35年的学术争论,指出真正与后者等效的是广义强度折减法,同时也指出,广义强度折减法的实用价值远不及理论价值,目前尚无有效的实‘现算法。从信息学的角度,严格地证明了斜坡临界滑动面搜索为NP完全问题(即：NP-Complete,简称NPC)。该证明的意义在于：首先,明确了不可能设计出具有多项式时间复杂度的精确求解算法,采用智能群体算法寻求近似解是解决问题的大方向；其次,从侧面证明基于二分法(含黄金分割法)的强度折减法与常规稳定性系数的定义方法并不完全等效。
(2)在提出二分法的并行多线程算法的基础上,将其应用于斜坡稳定性分析的强度折减法领域,提出了并行强度折减法。详细分析了并行算法的性能,给出了加速比的估算公式。对FLAC3D进行二次开发,将传统的FLAC3D环境下的强度折减法改造成并行的模式,效率提升显著,7个线程条件下的速度提高到原来的近4倍。
(3)提出了FLAC3D和有限元条件下的三维应力场(含物理指标参数场)的快速获取和访问算法。分层次、分阶段的逐步提高算法的效率,最终实现了应力场生成的多线程并行算法,大大提高效率,使得应力场生成和访问的时效性不再成为制约动力分析的瓶颈,为开展基于应力场的动力反应分析扫清了障碍。此外在地震波解译领域,开发了去除仪器响应的接口软件,填补了Windows下缺乏同类软件的空白。
(4)提出了斜坡非圆弧临界滑动面搜索的多层次并行微粒群算法。将该方法应用于斜坡动力反应分析,并结合谱分析理论,提出斜坡动力稳定系数的谱分析法。指出斜坡稳定系数具有明显的频率选择特性。将上述这些新方法引入高速远程滑坡的动力反应研究领域,以四川省北川县金鼓村滑坡为例,研究了动力甩出区的时间和空间演化规律,以及动力稳定系数的时域和频域特性。
斜坡动力反应分析是一个年轻的研究领域。目前研究还处于起步阶段,很多理论和方法尚不完善。本文仅作了初步的探索,笔者认为在以下几个方面仍然需要作进一步的研究：
(1)本文所建立的基于应力场的分析方法是以土坡为例的,对岩质斜坡问题如何引入此方法是需要进一步研究的问题。解决问题的方向有两条：一是沿着“岩体等效力学参数”的思路,对裂隙岩体采取等效处理,可以对某一类拓扑结构的岩体得到有一定共性的结论。其二是针对具体的岩质斜坡,运用工程地质的研究方法分析其可能存在的一组或几组潜在滑动面,并按结构面网络模拟的成果,将规模大的、有显著控制作用的优势面建立到模型中,以便采用特殊的接触面单元进行力学响应分析。无论采取上述何种方法,最终模型的应力场将能够在一定程度上反应裂隙岩体的力学特性,再通过本文所论述的滑动面随机搜索技术并结合搜索指定的潜在滑动面,便可综合研判最危险的情况。因此,沿用本文的基于应力场的处理思路将能够预期解决岩质斜坡的动力反应问题,其时间域、空间域、和频率域的分析方法是一致的。
(2)斜坡动力稳定性的可靠度研究。目前在岩土工程中,进行动力反应分析的岩土体参数的取值是个值得关注的问题。一般认为,数值模拟的输入参数不能直接采用室内实验参数,而是应该在工程地质类比的大框架下选取适当的值进行分析。因此,不同的学者取值就具有一定的随意性,如何反映不同的取值对最终稳定性评价的影响就需要进行可靠度研究。目前有关斜坡动力稳定性可靠度的研究成果还十分匮乏,这将是下一步研究的方向之一。
(3)继续将群体智能算法与斜坡滑移优化问题相结合,进行更深入的研究。本文提出的斜坡非圆弧临界滑动面搜索的多层次并行微粒群算法,预留了与其他优化方法融合的接口,下一步将寻求各种优化算法的融合,以便能够更好的解决实际问题。
(4)复杂条件下模型相似比的研究。在实际动力分析中发现,由于模型空间尺度太大,动辄上千米,即便大量的引入并行计算理论和方法,也无法完全摆脱运算效能不足的问题。解决问题的方向有两个,一是研究模型的相似比问题,即是希望通过某种小尺度的模型配以合适的参数来替代真实的模型,可以借鉴流体力学中无量纲化的思想,但在动力问题中由于涉及时间尺度,再加之很多分析软件并不支持完全无量纲化的分析,使得实现起来具有一定的困难,下一步应该加强此方面的研究。
(5)岩土力学领域高性能计算的研究。解决计算能力不足的另一条途径是继续加强高性能计算的研究,目前高性能计算领域的前沿有：多线程并行计算、基于GPU (Graphic Processing Unit)的并行计算、分布式集群计算和云计算,其中实现云计算具有极大的性能拓展空间。下一步的研究方向之一是全面改写已有的程序代码,使之能够解决三维复杂地层条件下的动力反应分析,包括滑动面搜索,时域和频域分析等子问题。
Slope is a kind of inclined geologic body with certain surface angle against horizon, which could be natural or human-made. The definition of slope contains both landslide and side slope Landslide is a kind of particular serious natural hazard, and among all kinds of slope hazard, landslide is the most dangerous one. Specifically, landslide trigged by earthquake can cause more severe hazard. Landslide triggered by Wenchuan earthquake is the major geo-hazard type in this area. The research of landslide triggered by earthquake requires the researchers to combine serial discipline together:engineering geology, hydraulic geology, constructive geology, earthquake engineering, rock and soil mechanics, etc.
This issue is written to analyze the dynamic behavior of the slope under seismic load. This issue is about to research the stability issues of certain type of slope, for the benefit of human beings.
The technical routine of this paper is to research the problem of general slope under equilibrium condition, and broad the result to the field of slope dynamic analysis and landslide dynamic analysis. The details of the technical routine are as follow:
(1) Based on the basic evaluating methods of the safety of slope, this issue defines the coefficient of stability in three ways:the basic definition of coefficient of stability, and definition based on the integration of the stress of sliding plane, and the definition of the reduction coefficient. The relationship among these define ways is carefully analyzed. The basic definition of coefficient of stability, and definition based on the integration of the stress of sliding plane is deeply relevant, the former is the advanced interpretation of the latter on the large scale. The definition of integration is the macroscopic definition compared to the basic one, both of them are defined under the reorganization of the unchanging state of stress and the material is viewed as frictional, and the coefficient is defined as the division of resisting stress to the frictional stress. And the reduction coefficient is based on the theory of finding the point of destruction, and defines the coefficient by reducing the strength. In the simple block conditions, strength reduction definition and the basic definition of law is equivalent, but in a complex stress field conditions, the situation is very different. To reveal the strength reduction coefficient with two different definitions, this paper explores the rock mass stress field and the physical and mechanical properties of the link between fields and then presented and proved the lemma of overall limited equilibrium state. Strictly through the lemma, the issue gives the prove that the mechanics and the golden section method based on the dichotomy of strength reduction and the definition of sliding surface stress integration is not equivalent, the equivalence is true with generalized strength reduction coefficient. The author defined that the so-called generalized strength reduction method is the stress field in the points with different reduction factors, the existence of an equivalent overall strength reduction factor is proved by the integral mean value theorem, which reflects the overall strength. This definition of coefficient is completely equivalent to the definition of coefficient based on the sliding surface stress integration. However, the theoretical value of the theory of generalized strength reduction is much larger than the practical value; there is no effective implementation algorithm for it.
(2) Discussed the applicability of strength reduction method. The usual bisection method provides a single reduction factor of strength, which is easily to utilize but not guaranteed. Although in most cases, the approximation is good, but it does not rule out the possibility of deviation under certain condition, especially when the geometry of the model is complicated. Numerical analysis shows the result of partial plastic flow, different from the whole stability we usually concerned. With this respect, the research on the circular arc sliding plane searching method under the definition of sliding plane stress integration is still needed for practical concern. By using the branch prediction, simultaneous multi-threading algorithm of bisection method is proposed, and applied to strength reduction calculation. A detailed analysis of the performance of parallel algorithms is presented, and the estimation formula of acceleration ratio is provided. Explored external plug-in for FLAC3D secondary development, changed the traditional FLAC3D environment strength reduction into parallel mode, which significantly improved the efficiency of calculation, gained nearly 4 times of acceleration when calculation with 7 multiple parallel threads.
(3)From view of informatics, concluded the question of slide plane searching, strictly proved that the searching for sliding plane of slope is one of the NPC type problem. This proved that is impossible to design the accurate algorithm with polynomial time complexity. Just for the good result of solving NPC problem as large groups of intelligent algorithms, using intelligent algorithms to find approximate solutions group to solve the problem of direction is applicable. Practice shows that, using particle swarm optimization (PSO) could obtain approximate solutions with good performance. The significance of that is, from the science of information it confirms that the dichotomy and the golden section method based on the strength reduction with the conventional definition of stability factor approach is not entirely equivalent, which fixed the argue of this issue.
(4) Studied non-circular critical slip surface searching method based on particle swarm optimization (PSO). The method uses sliding surface stress integration method to define the slope stability factor, and uses particle swarm optimization (PSO) to dynamically search the critical slip surface. The particle swarm optimization (PSO) applied to the critical slip surface searching method and implementation steps are completely stated. The multi-level particle swarm algorithm (PSO) is proposed, which established the foundation of combining the PSO algorithm with other optimization methods. Extend the PSO algorithm to parallel processing, which greatly increased the efficiency of slip surface searching, gained nearly 4 times of acceleration when calculation with 8 multiple parallel threads. Proposed the sliding surface dynamic random generation technology. In certain sense, the difficulty of introducing the particle swarm algorithm into a specific area, is not the particle swarm algorithm itself, but on how to associate the specific issues with the framework of algorithms. For the searching of sliding plane, the key factor for successfully utilizing PSO lies on the method of generating sliding plane with geometrical boundaries randomly. In this paper, for the example Australian Computer Aided Design Society (ACADS) examination questions EX1C, elaborated on randomly generated dynamic sliding surface technology, and on to program high-quality code under the specific situation..
(5) Presented the principle of velocity-type seismometer, details of the algorithm for removing instrument response, and developed the relevant interface software, which fills the blank of similar software for Windows. The software can automatically read the State Seismological Bureau of the SAC format, using time-domain filtering algorithm to remove instrument response, and automatically output speed and acceleration time series with the certain format which can be read directly for SeismoSignal software for further analysis. It is noteworthy that the main shock wave in Wenchuan is only speed-type seismograph records, but the speed type seismometer can not directly record the signal of acceleration process, in order to get the acceleration from the velocity record, numerical differentiation algorithm is demanded, the problems faced by numerical differentiation is the amplification of the data error. This is commonly referred to as the tendency of instability of differential. This paper describes and compares the accuracy of different algorithms for division of differential, pointing out that calculation is essentially equivalent to delay filtering, the higher the order of the more obvious of this equivalence, and in the contrary to the usual process of "smoothing", the acceleration calculation must reserve the minute information of the record. As well, the severe seismic condition of Wenchuan, it is not enough to record by speed type seismometer, instead, more accurate record requires strong seismometer with bigger measuring range.
(6) discussed in detail the solving process of dynamic problem of the slope with FLAC3D. Discussed in detail the boundary conditions, mechanical damping parameter selection method, mesh size and dynamic load matching filter choosing problem. Based on the record of Chendu(CD2), the EX1C example and the combination of vertical and horizontal seismic waves loading is applied to research on the dynamic behavior of models. Different from the regular one type seismic wave loading, this paper applied both vertical and horizontal waves, which are directed in EW and NS, separated in horizontal and vertical, so the true behavior is approach to the best extent. By using the parallel particle swarm algorithm of the slope of non-circular critical slip surface searching, obtain the position of most dangerous sliding plane at different time, and the corresponding record of stability coefficient versus time. By analyzing the stability of the coefficient curves of the frequency characteristics of the slope stability factor that has obvious frequency selectivity, prove that the frequency response of the dynamic load is uneven. Specifically for EX1C example, it is more sensitive to the low-frequency dynamic loads of 0.8HZ-3.2HZ.
(7) Developed a FLAC3D three-dimensional stress field conditions (including physical indicators parameter field) algorithm for fast access. Taking the directly to indirectly, space for time, serial to parallel as guiding ideology, increase the efficiency of the algorithm hierarchically and gradually. Eventually the paper completed the stress field generation in the multi-threaded parallel algorithm, which has produced extraordinary efficiency, high-speed access technology laid the foundation for the further stress field for complex conditions to a big degree of freedom dynamic analysis of slopes, making stress field generated and accessible time-constrain is no longer a bottleneck for dynamic response analysis. The author also developed a finite-element-method-based three-dimensional stress field conditions (including physical indicators parameter field) algorithm for fast access. Based on finite difference methods, most of the parameters of finite element stored in the grid node index, the situation should be more complicated. In this paper, an algorithm based on the volume weight to achieve the three-dimensional stress field under the conditions of the rapid access is applied, and inherited the FLAC3D rapid stress condition generation technology, finally accomplished the multi-threaded parallel computing.
(8) combined the basic method of slope dynamic analysis and the technology of non-arc sliding plane searching for slope, and utilize them in the research of dynamic behavior of slope. Take the Jingucun landslide of Beichuan City, Sichuan Province as example, this paper discussed the laws of dynamic pull-out area in time dimension, and the frequency characteristics of stability factor of the slope.
The innovations of this paper are as follow:
(1) Strictly through the lemma, the issue gives the prove that the mechanics and the golden section method based on the dichotomy of strength reduction and the definition of sliding surface stress integration is not equivalent, the equivalence is true with generalized strength reduction coefficient, made a closer of the 35-year long dispute in the rock mechanics field. It is pointed that the broad strength reduction method is more evaluable with its theoretical respect, there is no sufficient algorithm for it now. With the respect of informatics, strictly proved that the searching for sliding plane of slope is one of the NPC type problems.. The significance of the proof are:First, it is impossible to design a polynomial time complexity with the exact solution algorithm, using intelligent algorithms to find approximate solutions group to solve the problem; Secondly, the dichotomy-based (including golden section method) of strength reduction with conventional definitions of stability factor approach is not completely equivalent.
(2) Invented the parallel-thread strength reduction method. A detailed analysis of the performance of parallel algorithms is presented, and the estimation formula of acceleration ratio is provided. Explored external plug-in for FLAC3D secondary development, changed the traditional FLAC3D environment strength reduction into parallel mode, which significantly improved the efficiency of calculation, gained nearly 4 times of acceleration when calculation with 7 multiple parallel threads.
(3) Developed a FLAC3D three-dimensional stress field conditions (including physical indicators parameter field) algorithm for fast access. Improve the efficiency of the algorithm, eventually make possible the stress field generated by multi-threaded parallel algorithm, greatly improving efficiency, made it no longer the bottleneck of the dynamic analysis, which clean the barrier of dynamic analysis based on the stress field generation. In addition, in the field of seismic interpretation, developed the interface software to remove instrument response, which fill the blank of similar software for Windows.
(4) Studied non-circular critical slip surface searching method based on particle swarm optimization (PSO). The method is applied to dynamic analysis of slopes, combined with spectral analysis theory, it is proved that the slope stability factor has obvious characteristics of frequency selectivity. Combined the basic method of slope dynamic analysis and the technology of non-arc sliding plane searching for slope, and utilize them in the research of dynamic behavior of slope. Take the Jingucun landslide of Beichuan City, Sichuan Province as example, this paper discussed the laws of dynamic pull-out area in time dimension, and the frequency characteristics of stability factor of the slope.
Dynamic analysis of slope is a young field of research, many of the basic theory of slope stability analysis needs to learn, absorb and integrate. Current research still in its infancy, many theories and methods is not perfect. This is only a preliminary exploration, the following areas still need further research:
(1)The method brought in this thesis is specifically for soil slope, the comparative method for rock slope demands further research. To do so, two possible ways should be carefully developed:
One is to take the equivalent treatment of the fractured rock mass following the idea of 'equivalent parameters of rock mechanics', which could make the usage of uniform parameters possible for certain rock mass with similar topological structure. And the other one is to use of engineering geological research method to analyze the possible existence of a group or several groups of potential sliding surface for specific rock slope, according to network simulation results, add the controlling surface into the model established in order to use a special contact element analysis of the mechanical response. Both of the two way above are about to reflect the mechanical characteristics of fractured rock mass to certain extent, and then search the potential sliding plane using the method brought by this thesis. Therefore, it is participated that the dynamic response of rock slope can be fixed by using the method brought by this thesis, and its time domain, space domain, and frequency domain analysis method is consistent.
(2) reliability of Dynamic stability of slope. Currently in geotechnical engineering, the dynamic response analysis of rock and soil parameter value is a matter of concern. It is generally believed that laboratory parameters can not be directly used as the numerical simulation input parameters, but should select the appropriate value for analysis under the broad framework of the geological analogy. In this way, different scholars will arbitrary value, how to prove the impact of on different values on the final evaluation of the stability needs for reliability studies. At present the research of the reliability of dynamic stability of slope is very limited, which will be the further research directions.
(3) Continue to manage to combine the swarm intelligence algorithms with the research field of slope, and do further research on that. In this paper, the multi-level parallel particle swarm optimization of non-circular critical slip surface search of the slope allow integration with optimizations, the next step will be researching on the association of other optimizing algorism in order to solve and actual problems more properly and more accurately.
(4) Research of the ratio of similitude of model under complex conditions. Dynamic analysis found that in practical use, due to the spatial scale, often on the hundreds of kilometers, the short of efficiency of calculation is inevitable no matter how carefully the program has been written or how many the parallel analysis threads have been utilized. There are two possible ways to solve this problem:one is to research on the ratio of similitude, in other words, to use small scale models with proper parameters to preclude the full scale model; and the utilization of the philosophy of dimensionlessness of fluid mechanics. However, because of the time dimension in dynamic analysis, and the unsupportiveness of main software, the last way is hard to achieve., the next step should be to research in this area.
(5) Research of High Performance Computing method of Geotechnical engineering. Another way to solve the problem of deficiency of calculating ability is to invent more powerful calculation method, and there some several directions for this endeavor:parallel calculating, calculating based on GPU(Graphic Processing Unit), distributed cluster computing and cloud computing, among all those directions the cloud computing is most powerful. The next step is to rewrite all program code, to make the program appropriate to do dynamic analysis under the three-dimension complex geological conditions, including searching the sliding plane, time, and frequency analysis.
本文以地震条件下的斜坡为研究对象,从动力反应分析这个关键线路进行研究。目的是揭示强震条件下某一类斜坡岩土体稳定性评价问题,服务于生产,造福于人民。
本文采取的总体技术路线是,从一般斜坡静力学问题入手,逐步拓展到斜坡动力分析领域。全文的研究内容如下：
(1)从斜坡稳定性评价方法入手,在前人研究成果的基础上,将斜坡稳定系数的定义分为三大类,即：稳定系数的基本定义,稳定系数的滑动面应力积分定义和稳定系数的强度折减定义,研究三大类稳定系数的内在关系和差别。稳定系数的滑动面应力积分定义与稳定系数的基本定义关系密切,前者是后者在宏观尺度上的自然延伸,定义的内涵是相似的。这两种定义都是在不改变当前应力状态的情况下,按摩擦型材料的观点,以抗剪力与实际剪切应力的比值来定义稳定系数。而强度折减定义,本质上是寻找斜坡的临界破坏状态,以折减系数来描述强度储备,在解决问题的思路上明显不同。在简单块体条件下,强度折减法与基本定义法是等效的,但在复杂的应力场条件下,情况则有很大的不同。为了揭示强度折减法与前两类定义的差别,本文首先探讨应力场与岩土体物理力学性质场之间的联系,继而提出并证明了整体极限平衡状态引理。通过引理从力学上严格证明了基于二分法(含黄金分割法)的强度折减法与滑动面应力积分的定义并不完全等效,真正与之等效的是广义强度折减法。笔者定义的所谓广义强度折减法是指应力场中各点可具有不同的折减系数,由积分中值定理能够证明整体上存在一个等效的强度折减系数,该系数反映的是所有强度折减系数的加权平均,而且该系数与基于滑动面应力积分的定义完全等效。但广义强度折减法的实用价值远不及理论价值,目前尚无有效的实现算法。
(2)研究了强度折减法的适用性,并提出了并行强度折减法。通常基于二分法的强度折减法都是单一折减系数的强度折减,虽然简单易行,但是其等效性不能获得保证。虽然大多数情况,其逼近的特性良好,但不排除在某些情况下很可能会有较大的偏差,特别是在地质条件、几何形态复杂的条件下,各种数值模拟方法捕捉到的临界状态往往是局部的塑性流动,而不是我们所关心的整体极限平衡。因此,在滑动面应力积分定义的框架下开展非圆弧临界滑动面的搜索研究具有非常现实的意义。运用分支预测的思想,提出了二分法的并行多线程算法,并将其应用于强度折减法。详细分析了并行算法的性能,给出了加速比的估算公式。对FLAC3D进行二次开发,将传统的FLAC3D环境下的强度折减法改造成并行的模式,效率提升显著,7个线程条件下的速度提高到原来的近4倍。
(3)从信息学的角度,对滑动面搜索问题进行提炼,严格地证明了斜坡临界滑动面搜索为NPC类问题。证明的意义在于,首先明确了不可能设计出具有多项式时间复杂度的精确求解算法,为滑动面搜索算法的设计指明了方向：正如大量的NPC类问题采用群体智能算法取得很好的效果,采用智能群体算法寻求最优滑动面的近似解是解决问题的大方向。实践证明,采用微粒群算法(PSO)能够获得性能良好的近似解。证明的意义还在于,从信息学的角度印证了基于二分法(含黄金分割法)的强度折减法与常规稳定性系数的定义方法并不完全等效。
(4)研究了基于微粒群算法(PSO)的斜坡非圆弧临界滑动面的搜索方法。以滑动面应力积分的方法来定义斜坡稳定系数,并采用微粒群算法(PSO)来动态搜索临界滑动面。详细阐述了将微粒群算法(PSO)应用于临界滑动面搜索的方法和实现步骤。在此基础上提出了多层次微粒群算法(PSO),为PSO算法与其他优化方法融合奠定了基础。将PSO算法推广到并行处理,大幅度提高了滑动面搜索的效率,实测表明在8线程条件下能实现超过4倍的加速比。此外还详细研究了滑动面的动态随机生成技术。从某种意义上说,将微粒群算法引入一个具体的领域,其难度不在于微粒群算法本身,而在于如何将需要解决的具体问题与算法的框架模型相联系。具体到临界滑动面搜索问题,如何动态地随机生成符合几何约束条件的滑动面是微粒群算法(PSO)能否成功应用的关键。本文以澳大利亚计算机应用协会(ACADS)考核题EX1C为例,详细阐述了滑动面动态随机生成技术、对异常状况的处理方法,以及如何根据具体情况分析,编写高质量的代码,提高效率。
(5)开发了去除仪器响应的接口软件,填补了Windows下缺乏同类软件的空白。该软件能实现自动读取国家地震局公布的SEED/SAC格式文件,运用时域滤波算法去除仪器响应,并以文件形式自动输出速度和加速度时序,可直接供SeismoSignal软件读取进行后续分析。值得注意的是,关于汶川主震波形,目前可供使用的只有速度型地震仪的记录,但速度型地震仪并不能直接记录加速度时程信号,为了从速度时程得到加速度时程,需要采用数值微分算法,所面临的问题是数值微分往往会放大数据中的误差。这就是通常所说的“微分倾向于不稳定”。本文详细阐述和比较了不同的微商算法的精度问题,指出求解实际上等效于延时滤波,阶次越高越明显,与一般问题中倾向于“光滑”的处理恰好相反,加速度的求解应尽可能保留加速度时程的分辨率。同时指出,汶川强震条件下,速度型地震仪饱和现象十分明显,若要获取更为精确的波形则需要采用强震仪进行地震监测。
(6)研究了斜坡动力计算的边界条件的设置、力学阻尼参数的选取方法、网格尺寸和动力荷载滤波的匹配问题。以国家地震局成都台(CD2)记录的波形为动荷载,考察了ACADS考核题EX1C算例在纵、横波双通道耦合输入条件下的动力响应。与常规的岩土动力反应分析中仅仅使用一个通道的波形不同,本文采用纵波和横波耦合输入的方式,能够考虑东西向(EW)、南北向(NS)和垂直方向(H)三个方向上的耦合,尽可能真实的还原研究对象在真实地震条件下的反应。结合斜坡非圆弧临界滑动面搜索的多层次并行微粒群算法,获取了不同时最危险滑动面的位置,以及相应的稳定性系数时程曲线。通过谱分析,指出斜坡稳定系数对动力荷载频率的响应是不均匀的,具有明显的频率选择特性。具体到EX1C算例,对0.8HZ-3.2HZ的范围内的低频动力荷载比较敏感。
(7)开发了FLAC3D和有限元条件下的三维应力场(含物理指标参数场)的快速获取和访问算法。对于复杂的数值模拟问题,速度往往决定一切,否则我们就永远只能停留在解决简单的、概念性的教学模型阶段。与传统的静力分析不同,动力条件下,为了满足频谱分析的采样密度,对应力场获取的时效性提出了更高的要求,以往对此鲜有研究,如何短时间内快速的获取应力场是必须着力解决的问题。为此,笔者以变直接为间接、以空间换时间、改串行为并行的指导思想,分层次、分阶段的逐步提高算法的效率。最终实现了应力场生成和访问的多线程并行算法,获得了显著的效率提升。高速的应力场访问技术为深入进行复杂条件下的大自由度斜坡动力反应分析奠定了基础,使得应力场生成和访问的时效性不再成为制约动力分析的瓶颈,为开展基于应力场的动力反应分析扫清了障碍。有限元条件下与基于有限差分的方法不同,有限元的大部分参数指标保存在网格节点中,情况要更加复杂,本文采用基于体积权重的算法来实现三维条件下应力场的快速获取和访问,并继承了FLAC3D条件下的应力场快速生成技术,实现了多线程并行计算。
(8)把斜坡动力反应分析的基本方法、斜坡非圆弧临界滑动面搜索技术相结合,应用于高速远程滑坡的动力反应研究。以四川省北川县金鼓村滑坡为例,讨论了动力甩出区的时空演化规律,并分析了稳定系数、坡脚和坡顶等各监测点的加速度和速度的频率特性。
本文的主要创新点是：
(1)严格证明了基于二分法(含黄金分割法)的强度折减法与滑动面应力积分的定义并不完全等效,结束了岩土力学界长达35年的学术争论,指出真正与后者等效的是广义强度折减法,同时也指出,广义强度折减法的实用价值远不及理论价值,目前尚无有效的实‘现算法。从信息学的角度,严格地证明了斜坡临界滑动面搜索为NP完全问题(即：NP-Complete,简称NPC)。该证明的意义在于：首先,明确了不可能设计出具有多项式时间复杂度的精确求解算法,采用智能群体算法寻求近似解是解决问题的大方向；其次,从侧面证明基于二分法(含黄金分割法)的强度折减法与常规稳定性系数的定义方法并不完全等效。
(2)在提出二分法的并行多线程算法的基础上,将其应用于斜坡稳定性分析的强度折减法领域,提出了并行强度折减法。详细分析了并行算法的性能,给出了加速比的估算公式。对FLAC3D进行二次开发,将传统的FLAC3D环境下的强度折减法改造成并行的模式,效率提升显著,7个线程条件下的速度提高到原来的近4倍。
(3)提出了FLAC3D和有限元条件下的三维应力场(含物理指标参数场)的快速获取和访问算法。分层次、分阶段的逐步提高算法的效率,最终实现了应力场生成的多线程并行算法,大大提高效率,使得应力场生成和访问的时效性不再成为制约动力分析的瓶颈,为开展基于应力场的动力反应分析扫清了障碍。此外在地震波解译领域,开发了去除仪器响应的接口软件,填补了Windows下缺乏同类软件的空白。
(4)提出了斜坡非圆弧临界滑动面搜索的多层次并行微粒群算法。将该方法应用于斜坡动力反应分析,并结合谱分析理论,提出斜坡动力稳定系数的谱分析法。指出斜坡稳定系数具有明显的频率选择特性。将上述这些新方法引入高速远程滑坡的动力反应研究领域,以四川省北川县金鼓村滑坡为例,研究了动力甩出区的时间和空间演化规律,以及动力稳定系数的时域和频域特性。
斜坡动力反应分析是一个年轻的研究领域。目前研究还处于起步阶段,很多理论和方法尚不完善。本文仅作了初步的探索,笔者认为在以下几个方面仍然需要作进一步的研究：
(1)本文所建立的基于应力场的分析方法是以土坡为例的,对岩质斜坡问题如何引入此方法是需要进一步研究的问题。解决问题的方向有两条：一是沿着“岩体等效力学参数”的思路,对裂隙岩体采取等效处理,可以对某一类拓扑结构的岩体得到有一定共性的结论。其二是针对具体的岩质斜坡,运用工程地质的研究方法分析其可能存在的一组或几组潜在滑动面,并按结构面网络模拟的成果,将规模大的、有显著控制作用的优势面建立到模型中,以便采用特殊的接触面单元进行力学响应分析。无论采取上述何种方法,最终模型的应力场将能够在一定程度上反应裂隙岩体的力学特性,再通过本文所论述的滑动面随机搜索技术并结合搜索指定的潜在滑动面,便可综合研判最危险的情况。因此,沿用本文的基于应力场的处理思路将能够预期解决岩质斜坡的动力反应问题,其时间域、空间域、和频率域的分析方法是一致的。
(2)斜坡动力稳定性的可靠度研究。目前在岩土工程中,进行动力反应分析的岩土体参数的取值是个值得关注的问题。一般认为,数值模拟的输入参数不能直接采用室内实验参数,而是应该在工程地质类比的大框架下选取适当的值进行分析。因此,不同的学者取值就具有一定的随意性,如何反映不同的取值对最终稳定性评价的影响就需要进行可靠度研究。目前有关斜坡动力稳定性可靠度的研究成果还十分匮乏,这将是下一步研究的方向之一。
(3)继续将群体智能算法与斜坡滑移优化问题相结合,进行更深入的研究。本文提出的斜坡非圆弧临界滑动面搜索的多层次并行微粒群算法,预留了与其他优化方法融合的接口,下一步将寻求各种优化算法的融合,以便能够更好的解决实际问题。
(4)复杂条件下模型相似比的研究。在实际动力分析中发现,由于模型空间尺度太大,动辄上千米,即便大量的引入并行计算理论和方法,也无法完全摆脱运算效能不足的问题。解决问题的方向有两个,一是研究模型的相似比问题,即是希望通过某种小尺度的模型配以合适的参数来替代真实的模型,可以借鉴流体力学中无量纲化的思想,但在动力问题中由于涉及时间尺度,再加之很多分析软件并不支持完全无量纲化的分析,使得实现起来具有一定的困难,下一步应该加强此方面的研究。
(5)岩土力学领域高性能计算的研究。解决计算能力不足的另一条途径是继续加强高性能计算的研究,目前高性能计算领域的前沿有：多线程并行计算、基于GPU (Graphic Processing Unit)的并行计算、分布式集群计算和云计算,其中实现云计算具有极大的性能拓展空间。下一步的研究方向之一是全面改写已有的程序代码,使之能够解决三维复杂地层条件下的动力反应分析,包括滑动面搜索,时域和频域分析等子问题。
Slope is a kind of inclined geologic body with certain surface angle against horizon, which could be natural or human-made. The definition of slope contains both landslide and side slope Landslide is a kind of particular serious natural hazard, and among all kinds of slope hazard, landslide is the most dangerous one. Specifically, landslide trigged by earthquake can cause more severe hazard. Landslide triggered by Wenchuan earthquake is the major geo-hazard type in this area. The research of landslide triggered by earthquake requires the researchers to combine serial discipline together:engineering geology, hydraulic geology, constructive geology, earthquake engineering, rock and soil mechanics, etc.
This issue is written to analyze the dynamic behavior of the slope under seismic load. This issue is about to research the stability issues of certain type of slope, for the benefit of human beings.
The technical routine of this paper is to research the problem of general slope under equilibrium condition, and broad the result to the field of slope dynamic analysis and landslide dynamic analysis. The details of the technical routine are as follow:
(1) Based on the basic evaluating methods of the safety of slope, this issue defines the coefficient of stability in three ways:the basic definition of coefficient of stability, and definition based on the integration of the stress of sliding plane, and the definition of the reduction coefficient. The relationship among these define ways is carefully analyzed. The basic definition of coefficient of stability, and definition based on the integration of the stress of sliding plane is deeply relevant, the former is the advanced interpretation of the latter on the large scale. The definition of integration is the macroscopic definition compared to the basic one, both of them are defined under the reorganization of the unchanging state of stress and the material is viewed as frictional, and the coefficient is defined as the division of resisting stress to the frictional stress. And the reduction coefficient is based on the theory of finding the point of destruction, and defines the coefficient by reducing the strength. In the simple block conditions, strength reduction definition and the basic definition of law is equivalent, but in a complex stress field conditions, the situation is very different. To reveal the strength reduction coefficient with two different definitions, this paper explores the rock mass stress field and the physical and mechanical properties of the link between fields and then presented and proved the lemma of overall limited equilibrium state. Strictly through the lemma, the issue gives the prove that the mechanics and the golden section method based on the dichotomy of strength reduction and the definition of sliding surface stress integration is not equivalent, the equivalence is true with generalized strength reduction coefficient. The author defined that the so-called generalized strength reduction method is the stress field in the points with different reduction factors, the existence of an equivalent overall strength reduction factor is proved by the integral mean value theorem, which reflects the overall strength. This definition of coefficient is completely equivalent to the definition of coefficient based on the sliding surface stress integration. However, the theoretical value of the theory of generalized strength reduction is much larger than the practical value; there is no effective implementation algorithm for it.
(2) Discussed the applicability of strength reduction method. The usual bisection method provides a single reduction factor of strength, which is easily to utilize but not guaranteed. Although in most cases, the approximation is good, but it does not rule out the possibility of deviation under certain condition, especially when the geometry of the model is complicated. Numerical analysis shows the result of partial plastic flow, different from the whole stability we usually concerned. With this respect, the research on the circular arc sliding plane searching method under the definition of sliding plane stress integration is still needed for practical concern. By using the branch prediction, simultaneous multi-threading algorithm of bisection method is proposed, and applied to strength reduction calculation. A detailed analysis of the performance of parallel algorithms is presented, and the estimation formula of acceleration ratio is provided. Explored external plug-in for FLAC3D secondary development, changed the traditional FLAC3D environment strength reduction into parallel mode, which significantly improved the efficiency of calculation, gained nearly 4 times of acceleration when calculation with 7 multiple parallel threads.
(3)From view of informatics, concluded the question of slide plane searching, strictly proved that the searching for sliding plane of slope is one of the NPC type problem. This proved that is impossible to design the accurate algorithm with polynomial time complexity. Just for the good result of solving NPC problem as large groups of intelligent algorithms, using intelligent algorithms to find approximate solutions group to solve the problem of direction is applicable. Practice shows that, using particle swarm optimization (PSO) could obtain approximate solutions with good performance. The significance of that is, from the science of information it confirms that the dichotomy and the golden section method based on the strength reduction with the conventional definition of stability factor approach is not entirely equivalent, which fixed the argue of this issue.
(4) Studied non-circular critical slip surface searching method based on particle swarm optimization (PSO). The method uses sliding surface stress integration method to define the slope stability factor, and uses particle swarm optimization (PSO) to dynamically search the critical slip surface. The particle swarm optimization (PSO) applied to the critical slip surface searching method and implementation steps are completely stated. The multi-level particle swarm algorithm (PSO) is proposed, which established the foundation of combining the PSO algorithm with other optimization methods. Extend the PSO algorithm to parallel processing, which greatly increased the efficiency of slip surface searching, gained nearly 4 times of acceleration when calculation with 8 multiple parallel threads. Proposed the sliding surface dynamic random generation technology. In certain sense, the difficulty of introducing the particle swarm algorithm into a specific area, is not the particle swarm algorithm itself, but on how to associate the specific issues with the framework of algorithms. For the searching of sliding plane, the key factor for successfully utilizing PSO lies on the method of generating sliding plane with geometrical boundaries randomly. In this paper, for the example Australian Computer Aided Design Society (ACADS) examination questions EX1C, elaborated on randomly generated dynamic sliding surface technology, and on to program high-quality code under the specific situation..
(5) Presented the principle of velocity-type seismometer, details of the algorithm for removing instrument response, and developed the relevant interface software, which fills the blank of similar software for Windows. The software can automatically read the State Seismological Bureau of the SAC format, using time-domain filtering algorithm to remove instrument response, and automatically output speed and acceleration time series with the certain format which can be read directly for SeismoSignal software for further analysis. It is noteworthy that the main shock wave in Wenchuan is only speed-type seismograph records, but the speed type seismometer can not directly record the signal of acceleration process, in order to get the acceleration from the velocity record, numerical differentiation algorithm is demanded, the problems faced by numerical differentiation is the amplification of the data error. This is commonly referred to as the tendency of instability of differential. This paper describes and compares the accuracy of different algorithms for division of differential, pointing out that calculation is essentially equivalent to delay filtering, the higher the order of the more obvious of this equivalence, and in the contrary to the usual process of "smoothing", the acceleration calculation must reserve the minute information of the record. As well, the severe seismic condition of Wenchuan, it is not enough to record by speed type seismometer, instead, more accurate record requires strong seismometer with bigger measuring range.
(6) discussed in detail the solving process of dynamic problem of the slope with FLAC3D. Discussed in detail the boundary conditions, mechanical damping parameter selection method, mesh size and dynamic load matching filter choosing problem. Based on the record of Chendu(CD2), the EX1C example and the combination of vertical and horizontal seismic waves loading is applied to research on the dynamic behavior of models. Different from the regular one type seismic wave loading, this paper applied both vertical and horizontal waves, which are directed in EW and NS, separated in horizontal and vertical, so the true behavior is approach to the best extent. By using the parallel particle swarm algorithm of the slope of non-circular critical slip surface searching, obtain the position of most dangerous sliding plane at different time, and the corresponding record of stability coefficient versus time. By analyzing the stability of the coefficient curves of the frequency characteristics of the slope stability factor that has obvious frequency selectivity, prove that the frequency response of the dynamic load is uneven. Specifically for EX1C example, it is more sensitive to the low-frequency dynamic loads of 0.8HZ-3.2HZ.
(7) Developed a FLAC3D three-dimensional stress field conditions (including physical indicators parameter field) algorithm for fast access. Taking the directly to indirectly, space for time, serial to parallel as guiding ideology, increase the efficiency of the algorithm hierarchically and gradually. Eventually the paper completed the stress field generation in the multi-threaded parallel algorithm, which has produced extraordinary efficiency, high-speed access technology laid the foundation for the further stress field for complex conditions to a big degree of freedom dynamic analysis of slopes, making stress field generated and accessible time-constrain is no longer a bottleneck for dynamic response analysis. The author also developed a finite-element-method-based three-dimensional stress field conditions (including physical indicators parameter field) algorithm for fast access. Based on finite difference methods, most of the parameters of finite element stored in the grid node index, the situation should be more complicated. In this paper, an algorithm based on the volume weight to achieve the three-dimensional stress field under the conditions of the rapid access is applied, and inherited the FLAC3D rapid stress condition generation technology, finally accomplished the multi-threaded parallel computing.
(8) combined the basic method of slope dynamic analysis and the technology of non-arc sliding plane searching for slope, and utilize them in the research of dynamic behavior of slope. Take the Jingucun landslide of Beichuan City, Sichuan Province as example, this paper discussed the laws of dynamic pull-out area in time dimension, and the frequency characteristics of stability factor of the slope.
The innovations of this paper are as follow:
(1) Strictly through the lemma, the issue gives the prove that the mechanics and the golden section method based on the dichotomy of strength reduction and the definition of sliding surface stress integration is not equivalent, the equivalence is true with generalized strength reduction coefficient, made a closer of the 35-year long dispute in the rock mechanics field. It is pointed that the broad strength reduction method is more evaluable with its theoretical respect, there is no sufficient algorithm for it now. With the respect of informatics, strictly proved that the searching for sliding plane of slope is one of the NPC type problems.. The significance of the proof are:First, it is impossible to design a polynomial time complexity with the exact solution algorithm, using intelligent algorithms to find approximate solutions group to solve the problem; Secondly, the dichotomy-based (including golden section method) of strength reduction with conventional definitions of stability factor approach is not completely equivalent.
(2) Invented the parallel-thread strength reduction method. A detailed analysis of the performance of parallel algorithms is presented, and the estimation formula of acceleration ratio is provided. Explored external plug-in for FLAC3D secondary development, changed the traditional FLAC3D environment strength reduction into parallel mode, which significantly improved the efficiency of calculation, gained nearly 4 times of acceleration when calculation with 7 multiple parallel threads.
(3) Developed a FLAC3D three-dimensional stress field conditions (including physical indicators parameter field) algorithm for fast access. Improve the efficiency of the algorithm, eventually make possible the stress field generated by multi-threaded parallel algorithm, greatly improving efficiency, made it no longer the bottleneck of the dynamic analysis, which clean the barrier of dynamic analysis based on the stress field generation. In addition, in the field of seismic interpretation, developed the interface software to remove instrument response, which fill the blank of similar software for Windows.
(4) Studied non-circular critical slip surface searching method based on particle swarm optimization (PSO). The method is applied to dynamic analysis of slopes, combined with spectral analysis theory, it is proved that the slope stability factor has obvious characteristics of frequency selectivity. Combined the basic method of slope dynamic analysis and the technology of non-arc sliding plane searching for slope, and utilize them in the research of dynamic behavior of slope. Take the Jingucun landslide of Beichuan City, Sichuan Province as example, this paper discussed the laws of dynamic pull-out area in time dimension, and the frequency characteristics of stability factor of the slope.
Dynamic analysis of slope is a young field of research, many of the basic theory of slope stability analysis needs to learn, absorb and integrate. Current research still in its infancy, many theories and methods is not perfect. This is only a preliminary exploration, the following areas still need further research:
(1)The method brought in this thesis is specifically for soil slope, the comparative method for rock slope demands further research. To do so, two possible ways should be carefully developed:
One is to take the equivalent treatment of the fractured rock mass following the idea of 'equivalent parameters of rock mechanics', which could make the usage of uniform parameters possible for certain rock mass with similar topological structure. And the other one is to use of engineering geological research method to analyze the possible existence of a group or several groups of potential sliding surface for specific rock slope, according to network simulation results, add the controlling surface into the model established in order to use a special contact element analysis of the mechanical response. Both of the two way above are about to reflect the mechanical characteristics of fractured rock mass to certain extent, and then search the potential sliding plane using the method brought by this thesis. Therefore, it is participated that the dynamic response of rock slope can be fixed by using the method brought by this thesis, and its time domain, space domain, and frequency domain analysis method is consistent.
(2) reliability of Dynamic stability of slope. Currently in geotechnical engineering, the dynamic response analysis of rock and soil parameter value is a matter of concern. It is generally believed that laboratory parameters can not be directly used as the numerical simulation input parameters, but should select the appropriate value for analysis under the broad framework of the geological analogy. In this way, different scholars will arbitrary value, how to prove the impact of on different values on the final evaluation of the stability needs for reliability studies. At present the research of the reliability of dynamic stability of slope is very limited, which will be the further research directions.
(3) Continue to manage to combine the swarm intelligence algorithms with the research field of slope, and do further research on that. In this paper, the multi-level parallel particle swarm optimization of non-circular critical slip surface search of the slope allow integration with optimizations, the next step will be researching on the association of other optimizing algorism in order to solve and actual problems more properly and more accurately.
(4) Research of the ratio of similitude of model under complex conditions. Dynamic analysis found that in practical use, due to the spatial scale, often on the hundreds of kilometers, the short of efficiency of calculation is inevitable no matter how carefully the program has been written or how many the parallel analysis threads have been utilized. There are two possible ways to solve this problem:one is to research on the ratio of similitude, in other words, to use small scale models with proper parameters to preclude the full scale model; and the utilization of the philosophy of dimensionlessness of fluid mechanics. However, because of the time dimension in dynamic analysis, and the unsupportiveness of main software, the last way is hard to achieve., the next step should be to research in this area.
(5) Research of High Performance Computing method of Geotechnical engineering. Another way to solve the problem of deficiency of calculating ability is to invent more powerful calculation method, and there some several directions for this endeavor:parallel calculating, calculating based on GPU(Graphic Processing Unit), distributed cluster computing and cloud computing, among all those directions the cloud computing is most powerful. The next step is to rewrite all program code, to make the program appropriate to do dynamic analysis under the three-dimension complex geological conditions, including searching the sliding plane, time, and frequency analysis.
引文
[1]殷跃平.中国地质灾害减灾回顾与展望——从国际减灾十年到国际减灾战略[J].国土资源科技管理.2001(03)：26-29.
[2]殷跃平.汶川八级地震地质灾害研究[J].工程地质学报.2008(04)：433-444.
[3]殷跃平.汶川八级地震滑坡高速远程特征分析[J].工程地质学报.2009(02)：153-166.
[4]殷跃平.汶川八级地震滑坡特征分析[J].工程地质学报.2009(01)：29-38.
[5]Jibson R W, Harp E L, Michael J A. A method for producing digital probabilistic seismic landslide hazard maps[J]. Engineering Geology.2000,58(3/4):271-289.
[6]Keefer D V. Statistical analysis of an earthquake-induced landslidedistribution-the 1989 Loma Prieta, California event[J]. Engineering Geology.2000,58(3/4):231-249.
[7]Mahdavifar M R, Solaymani S, Jafari M K. Landslides triggered by the Avaj, Iran earthquake of June 22,2002[J]. Engineering Geology.2006,86(2/3):166-182.
[8]Wang W N, Nakamura H, Tsuchiya S, et al. Distributions of landslides triggered by the Chi-Chi earthquake in Central Taiwan on September 21,1999[J]. Journal of the Japan Landslide Society. 2002,38(4):18-26.
[9]晏同珍,杨顺安,方云.滑坡学[M].武汉：中国地质大学出版社,2000.
[10]Heim A. Bergsturz und Menschenleben[M]. Zutich:Naturforschenden Gesellschaft,1989.
[11]Scheidegger A E. On the prediction of the reach and velocity of catastrophic landslides [J]. Rock Mechanics and Rock Engineering.1973,5(4):231-236.
[12]Voight B. Rockslides and avalanches[M]. Amsterdam:Elsevier Scientific Pub. Co.,1978.
[13]Bock C G. Martinezmountain rock avalanche[C]. U.S.:Geological Society of America,1977. 155-168.
[14]Evans S G, Hungr O, Clague J J. Dynamics of the 19rock avalanche and associated distal debris flow onMountCayleBritish Columbia, Canada:Implications for landslide hazard sessment on dissected volcanoes[J]. Engineering Geology.2006,1(1):29-51.
[15]Qiao L, Yuan C, Miyajima M, et al. Shake-table testing and FLAC modeling of liquefaction-induced slope failure and damage to buried pipelines[C]. Sacramento, CA, United states:American Society of Civil Engineers,2008.
[16]Sun J, Tian X, Guan X, et al. Stability Analysis for Loosened Rock Slope of Jinyang Grand Buddha in Taiyuan, China[J]. Earth Science Frontiers.2008,15(4):227-238.
[17]He M C, Feng J L, Sun X M. Stability evaluation and optimal excavated design of rock slope at Antaibao open pit coal mine, China[J]. International Journal of Rock Mechanics and Mining Sciences.2008,45(3):289-302.
[18]Itasca. FLAC-3D fast Lagrangian analysis of continua in 3dimension version 3.0[R]. USA: Itasca ConsuItingGroup Inc.,2005.
[19]Huang X, Mao X, Miao X, et al. Study of slope reinforcement force based on FLAC[J]. Journal of China University of Mining and Technology.2006,16(2):123-127.
[20]Pan Q, Chang L. A study of application about FLAC-3D in damage of rock mass slope[J]. Key Engineering Materials.2006,306-308 II:1427-1432.
[21]崔芳鹏,胡瑞林,张明,等.重庆小南海地震崩滑体的基本特征及形成机制研究[J].工程地质学报.2009(02)：167-174.
[22]李维树,谭新,黄志鹏.长江小南海水利枢纽坝基弱风化岩体工程力学性质研究及可利用探讨[J].岩石力学与工程学报.2008(S2)：3899-3904.
[23]Kim D, Pan X, Pardalos P M. An Enhanced Dynamic Slope Scaling Procedure with Tabu Scheme for Fixed Charge Network Flow Problems[J]. Computational Economics.2006.
[24]Kim D, Pan X, Pardalos P. An Enhanced Dynamic Slope Scaling Procedure with Tabu Scheme for Fixed Charge Network Flow Problems[J]. Landslides.2006,27(2):273-293.
[25]刘红帅,薄景山,刘德东.岩土边坡地震稳定性分析研究评述[J].地震工程与工程振动.2005(01)：164-171.
[26]刘红帅,薄景山,刘德东.岩土边坡地震稳定性评价方法研究进展[J].防灾科技学院学报.2007(03)：20-27.
[27]祁生文,伍法权,刘春玲.岩质边坡动力反应分析[M].北京：科学出版社,2007.
[28]Lawphongpanich S. Dynamic Slope Scaling Procedure and Lagrangian Relaxation with Subproblem Approximation[J]. Doklady Earth Sciences.2006,35(1):121-130.
[29]Al-Homoud A S, Tahtamoni W W. Reliability analysis of three-dimensional dynamic slope stability and earthquake-induced permanent displacement[J]. Soil Dynamics and Earthquake Engineering.2000,19(2):91-114.
[30]Stewart J P, Blake T F, Hollingsworth R A. A Screen Analysis Procedure for Seismic Slope Stability[J]. Earthquake Spectra.2003,19(3):697-712.
[31]Lin M, Wang K. Seismic slope behavior in a large-scale shaking table model test[J]. Engineering Geology.2006,86(2-3):118-133.
[32]Urgeles R, Locat J, Lee H J, et al. The Saguenay Fjord, Quebec, Canada:integrating marine geotechnical and geophysical data for spatial seismic slope stability and hazard assessment[J]. Marine Geology.2002,185(3-4):319-340.
[33]Kokusho T, Ishizawa T, Nishida K. Travel distance of failed slopes during 2004 Chuetsu earthquake and its evaluation in terms of energy[J]. Soil Dynamics and Earthquake Engineering. 2009,29(7):1159-1169.
[34]唐辉明,滕伟福.岩土工程数值模拟新方法[J].地质科技情报.1998(S2)：42-49.
[35]De Santayana F P, Roque A J. Preliminary static and seismic slope stability analyses for a Portuguese municipal solid waste landfill[C]. Cardiff, Wales, United kingdom:Thomas Telford Services Ltd,2006.
[36]Urgeles R, Locat J, Lee H J, et al. The Saguenay Fjord, Quebec, Canada:Integrating marine geotechnical and geophysical data for spatial seismic slope stability and hazard assessment[J]. Marine Geology.2002,185(3-4):319-340.
[37]Bray J. Simplified Seismic Slope Displacement Procedures [J]. Geo-Marine Letters.2007: 327-353.
[38]Yegian M K, Harb J N. Slip displacements of geosynthetic systems underdynamic excitation[M]. Design and performance of solid waste landfills, No.54, Yegian M K, Finn W, San Diego, California, ASCE,:Geotechnical Special Publication,1995,212-236.
[39]Yegian M K, Lahlfa A M. Dynamic interface shear strength properties of geomembrabes and geotexiles[J]. J Geotech Engrg ASCE.1992,118(5):760-779.
[40]Terzaghi K. Mechanisms of landslides[C].1950.83-123.
[41]Seed H B. Considerations in the earthquake-resistant design of earth and rockfill dams[J]. Geotechnique.1979,29(3):215-263.
[42]Mononobe A H. Seismic stability of the earth dam[C]. Washinton DC:Ⅳ:1936.
[43]Seed H B, Martin R G. The seismic coefficient in earth dam design[J]. Journal of Soil mechanics and Foundation Division.1996,104(SM3):59-83.
[44]Newmark N M. Effect of earthquakes on dams and embankments [J]. Geotechnique.1965,15(2): 139-160.
[45]Wang K, Lin M. Development of shallow seismic landslide potential map based on Newmark's displacement:the case study of Chi-Chi earthquake, Taiwan[J]. Ecological Research.
[46]Ingles J, Darrozes J, Soula J. Effects of the vertical component of ground shaking on earthquake-induced landslide displacements using generalized Newmark analysis [J]. Engineering Geology.2006,86(2-3):134-147.
[47]Chang K, Taboada A, Lin M, et al. Analysis of landsliding by earthquake shaking using a block-on-slope thermo-mechanical model:Example of Jiufengershan landslide, central Taiwan[J]. Engineering Geology.2005,80(1-2):151-163.
[48]Lin J S, Whitman R V. Earthquake induced displacements of sliding blocks[J]. J Geotech Engng Div.1986,112(1):44-59.
[49]Makdisi F I, Seed H B. Simplified procedure for estimating dam and embankment earthquake induced deformations [J]. J Geotech Engng.1978,104(No.GT7):569-587.
[50]Martin R G, Finn W D L, Seed H B. Fundamentals of liquefaction under cyclic loading[J]. J Geotech Engrg Div asce.1975,101(GT5):423-438.
[51]Sarma S K. Seismic stability of earth dams and embankments[J]. Geotech Engng Div.1975,105, GT4:449-464.
[52]Sarma S K. Seismic displacement analysis of earth dams[J]. J Geotech Engng Div 107.1981, GT12:1735-1739.
[53]Wang S J, Xue S Y, Maugeri M. Dynamic stability of the left abutment in the Xiaolangdi Project On the Yellow River[C]. Istanbul(Turkey):A. a. Balma, Rotterdam,1993.
[54]薄景山,徐国栋,景立平.土边坡地震反应及其动力稳定性分析[J].地震工程与工程振动.2001(02)：116-120.
[55]陈玲玲,陈敏中,钱胜国.岩质陡高边坡地震动力稳定分析[J].长江科学院院报.2004(01)：33-35.
[56]Medina C L, Zeghal M. A micro-mechanical study of the dynamic response of unsaturated level sites and sloping grounds[C]. Denver, CO, United states:American Society of Civil Engineers, 2007.
[57]Lollino P, Santaloia F, Amorosi A, et al. Numerical analysis of the behaviour of a cut slope in stiff clays by finite difference method[C]. Graz, Austria:Taylor and Francis/Balkema,2006.
[58]Zou J Z, Willians D J, Xiong W L. Search for critical slip surfaces based on finite element method[J]. Canadian Geotechnical Journal.1995,32:233-246.
[59]Soares Jr. D. An optimised FEM-BEM time-domain iterative coupling algorithm for dynamic analyses[J]. Computers & Structures.2008,86(19-20):1839-1844.
[60]Wang X, Wang L B. Dynamic analysis of a water-soil-pore water coupling system[J]. Computers & Structures.,85(11-14):1020-1031.
[61]Assimaki D, Kausel E. Modified topographic amplification factors for a single-faced slope due to kinematic soil-structure interaction[J]. Journal of Geotechnical and Geoenvironmental Engineering.2007,133(11):1414-1431.
[62]Ford S A A S, Sitar N, Lysmer J, et al. Topographic effects on the seismic response of steep slopes[J]. Bulletin of the Seismological Society of America.1997,87(3):701-709.
[63]刘春玲,祁生文,童立强,等.利用FLAC-(3D)分析某边坡地震稳定性[J].岩石力学与工程学报.2004(16)：2730-2733.
[64]许瑾,郑书英.边界元法分析边坡动态稳定性[J].西北建筑工程学院学报(自然科学版).2000(04).
[65]孟达,孟祥冲,赵援.边坡的开挖及其地震动力响应分析[J].煤炭工程.2009(01)：77-80.
[66]杨扬.边坡稳定性分析方法简介[J].科技信息(科学教研).2008(03)：63.
[67]杨俊凯.边坡稳定性评价方法综述[J].中国科技信息.2009(16)：65-66.
[68]张云.露天矿边坡岩体裂隙含水介质渗流场边界元分析[J].南京大学学报数学半年刊.1992(01)：119-125.
[69]李建平,熊传治.岩石边坡安全系数的边界元解[J].岩石力学与工程学报.1990(03)：238-243.
[70]廖小平.一种新的边坡模拟技术[J].路基工程.1993(01)：43-47.
[71]樊文熙.有限元边界元法在岩体力学与工程中的应用[J].中州煤炭.1985(04)：2-3.
[72]Chunhan Z, Pekau O A, Feng J, et al. Application of disticnct element method in dynamic analysis of high rock slopes and blocky structures [J]. Soil Dynamics and Earthquake Engineering.1997,16(6):385-394.
[73]张建海,范景伟,何江达.用刚体弹簧元法求解边坡、坝基动力安全系数[J].岩石力学与工程学报.1999(04)：19-23.
[74]刘君,孔宪京,Shyu Kuokai. DDA与FEM耦合法在分缝重力坝非线性分析中的应用[J].计算力学学报.2004(05)：585-591.
[75]邓学晶,孔宪京,刘君.城市垃圾填埋场的地震响应及稳定性分析[J].岩土力学.2007,22(10)：2095-2100.
[76]刘君,孔宪京.卫生填埋场复合边坡地震稳定性和永久变形分析[J].岩土力学.2004,22(05)：778-782.
[77]卓家涛,章青.不连续介质力学问题的界面元法[M].北京：科学出版社,2004.
[78]裴觉民.数值流形方法与非连续变形分析[J].岩石力学与工程学报.1997(03)：80-93.
[79]Clough W R, Chopra K A. Earthquake stress analysis in earth dams[J]. J Engrg Mech, ASCE. 1966,92(2):197-211.
[80]谢康和,周建.岩土工程有限元分析理论与应用[M].北京：科学出版社,2000.
[81]李鸿波,张我华,陈云敏.爆炸冲击荷载作用下重力坝三维各向异性脆性动力损伤有限元分析[J].岩石力学与工程学报.2006(08).
[82]刘红帅,唐立强,薄景山,等.岩体弹粘塑性显式波动有限元分析[J].煤田地质与勘探.2009(01)：38-42.
[83]董志良,陈平山,莫海鸿,等.真空预压法有限元计算比较[J].岩石力学与工程学报.2008(11).
[84]熊敬,张建海Druker-Prager型屈服准则与强度储备安全系数的相关分析[J].岩土力学.2008,22(07)：1905-1910.
[85]郑宏,刘德富.弹塑性矩阵D_(ep)的特性和有限元边坡稳定性分析中的极限状态标准[J].岩石力学与工程学报.2005(07).
[86]许建聪,尚岳全,郑束宁,等.不平衡推力法的弹塑性有限元改进[J].岩石力学与工程学报.2005(23).
[87]王海斌,李永盛.边坡稳定性有限元分析的处理技巧[J].岩石力学与工程学报.2005(13).
[88]曾亚武,田伟明.边坡稳定性分析的有限元法与极限平衡法的结合[J].岩石力学与工程学报.2005(S2).
[89]孙世永.土动力学研究与发展综述[J].四川建筑.2006(03)：68-69.
[90]Cundall A P. A computer model for simulating progressive larg scale movement in blocky rock system[C]. Nancy France:1971.
[91]Itasca. Universal Distinct Element Code, Version 3.1 User's Manual[M]. Minneapolis, Minnesota USA:Itasca Consulting Group Inc.,1999.
[92]Bardet J, F S R. Seismic stability of fracture rock masses with the distinct element method[C]. Rapid City:1985.
[93]陶连金,苏生瑞,张倬元.节理岩体边坡的动力稳定性分析[J].工程地质学报.2001(01)：32-38.
[94]祁生文,伍法权,兰恒星.盘石头水库泄洪洞、导流洞进出口高边坡稳定性分析[J].岩石力学与工程学报.2002(03)：353-359.
[95]祁生文,伍法权,刘春玲,等.地震边坡稳定性的工程地质分析[J].岩石力学与工程学报. 2004(16)：2792-2797.
[96]祁生文,伍法权,孙进忠.边坡动力响应规律研究[J].中国科学E辑.2003(S1)：28-40.
[97]卢开澄,卢华明.组合数学[M].北京：清华大学出版社,2002.
[98]Xu Q, Fan X, Huang R, et al. Landslide dams triggered by the Wenchuan Earthquake, Sichuan Province, south west China[J]. Bulletin of Engineering Geology and the Environment.2009, 68(3):373-386.
[99]Tang C, Zhu J, Liang J. Emergency assessment of seismic landslide susceptibility:a case study of the 2008 Wenchuan earthquake affected area[J]. Acta Seismologica Sinica.2009,8(2): 207-217.
[100]Cui P, Zhu Y, Han Y, et al. The 12 May Wenchuan earthquake-induced landslide lakes: Distribution and preliminary risk evaluation[J]. Landslides.2009,6(3):209-223.
[101]王成华,夏绪勇.边坡稳定分析中的临界滑动面搜索方法评述[J].四川建筑科学研究院.2002,28(3).
[102]陈昌富,龚晓南.混沌扰动启发式蚁群算法及其在边坡非圆弧临界滑动面搜索中的应用[J].岩石力学与工程学报.2004(20)：3450-3453.
[103]Chen Y M. Location of critical failure surface and some further studies on slope stability analysis[J]. Computers and Geotechnics.2003,30(3):255-267.
[104]Zhu D Y. A method for locating critical slip surfaces in slope stability analysis[J]. Canadian Geotechnical Journal.2001,38:328-337.
[105]Mccombie P, Wilkinson P. The use of the simple genetic algorithm in finding the critical factor of safety in slope stability analysis[J]. Computers and Geotechnics.2002,29(8):699-714.
[106]Goh A T C. Genetic algorithm search for critical slip surface in multiple-wedge stability analysis[J]. Canadian Geotechnical Journal.1999,36:382-391.
[107]Boutrup E, Lovell C W. Search technique in slope stability analysis[J]. Engineering Geology. 1980,16(1):51-61.
[108]Siegel R A, Kovacs W D, Lovell C W. Random surface generation in stability analysis[J]. Journal of Geotechnical Engineering.1981,107(7):991-996.
[109]Baker R. Determination of the critical slip surface in slope stability computation[J]. International Journal for Numerical and Analytical Methods in Geomechanics.1980,4(4): 333-359.
[110]Pham H T V, Fredlund D G, Fredlund D G. The application of dynamic programming to slope stability analysis[J]. Canadian Geotechnical Journal.2003,40(830-847).
[111]曹文贵,颜荣贵.边坡非圆临界滑面确定之动态规划法研究[J].岩石力学与工程学报.1995(04)：320-328.
[112]陈云敏,魏新江,李育超.边坡非圆弧临界滑动面的粒子群优化算法[J].岩石力学与工程学报.2006(07)：1443-1449.
[113]李育超,凌道盛,陈云敏,等.蒙特卡洛法与有限元相结合分析边坡稳定性[J].岩石力学与工程学报.2005(11)：1933-1941.
[114]陈云敏,李育超,凌道盛.蒙特卡洛法与有限元结合搜索边坡临界滑动面[J].岩土力学.2004(S2)：75-80.
[115]Tingey B E, Mcbride J H,Thompson T J, et al. Study of a prehistoric landslide using seismic reflection methods integrated with geological data in the Wasatch Mountains, Utah, USA[J]. Engineering Geology.2007,95(1-2):1-29.
[116]Sutter H. The Free Lunch Is Over:A Fundamental Turn Toward Concurrency in Software[J]. Dr. Dobb's Journal.2005,30(3):16-20,22.
[117]张舒,褚艳利,赵开勇,等.GPU高性能运算之CUDA[M].北京：中国水利水电出版社,2009.
[118]维基百科编者.摩尔定律维基百科,2010,http://zh.wikipedia.org/w/index.php?title=%E6%91%A9%E5%B0%94%E5%AE%9A%E 5%BE%8B&oldid=12765148.
[119]Enet硅谷动力.英特尔80核心cpu研发成功2010,http://www.cz7v.com/news/zixun/201001/58.html.
[120]郑宏,田斌,刘德富,等.关于有限元边坡稳定性分析中安全系数的定义问题[J].岩石力学与工程学报.2005(13)：2225-2230.
[121]Zienkiewicz O C, Humpheson C, Lewis R W. Associated and non-associated visco-plasticity and plasticity in soil mechanics[J]. Geotechnique.1975,25(4):671-689.
[122]Lin H, Cao P, Li J, et al. Displacement characteristic of soil slope in shear strength reduction method based on Mohr-coulomb model[C]. China:Springer-Verlag GmbH and Co. KG,2008.
[123]Jiang Q. Strength reduction method for slope based on a ubiquitous-joint criterion and its application[J]. Mining Science and Technology (China).2009,19(4):452-456.
[124]刘金龙,栾茂田,赵少飞,等.关于强度折减有限元方法中边坡失稳判据的讨论[J].岩土力学.2005(08)：1345-1348.
[125]张鲁渝,郑颖人,赵尚毅.有限元强度折减系数法计算土坡稳定安全系数的精度研究[J].水利学报.2003(1)：21-26.
[126]张士兵,王练柱,张建.边坡稳定性弹塑性大变形有限元强度折减分析[J].岩石力学与工程学报.2004(S1)：4463-4467.
[127]赵尚毅,郑颖人,邓卫东.用有限元强度折减法进行节理岩质边坡稳定性分析[J].岩石力学与工程学报.2003(02)：254-260.
[128]郑颖人,赵尚毅,邓卫东.岩质边坡破坏机制有限元数值模拟分析[J].岩石力学与工程学报.2003(12)：1943-1952.
[129]杨有成.强度折减法在斜坡稳定性分析中的适用性研究[D].武汉：中国地质大学(武汉),2008.
[130]杨有成,李群,陈新泽,等.对强度折减法若干问题的讨论[J].岩土力学.2008,22(04)：1103-1106.
[131]陈新泽,唐辉明,杨有成,等.基于FLAC-(3D)强度折减法滑坡三维稳定性研究——以三峡库区白果树古滑坡群为例[J].水文地质工程地质.2008(02)：24-29.
[132]郑颖人,张玉芳,赵尚毅,等.有限元强度折减法在元磨高速公路高边坡工程中的应用[J].岩石力学与工程学报.2005(21).
[133]翟长海,谢礼立.结构抗震设计中的强度折减系数研究进展[J].哈尔滨工业大学学报.2007(08)：1177-1184.
[134]陈菲,邓建辉.岩坡稳定的三维强度折减法分析[J].岩石力学与工程学报.2006(12).
[135]段荣福,白世伟.基于临界滑动场的岩质边坡临界滑动面搜索方法[J].岩土力学.2007(05)：861-864.
[136]吴余生,陈胜宏.并行组合模拟退火算法在边坡稳定分析中的应用[J].岩土力学.2006(09)：1554-1558.
[137]李梦玄,余飞,王元汉.边坡失稳的仿真模拟分析[J].岩土力学.2002(S1)：169-171.
[138]陈祖煜.土质边坡稳定分析：原理·方法·程序[M].北京：中国水利水电出版社,2003.
[139]Intel.Intel微处理器出口规范测量机制,2010,http://www.intel.com/support/tw/processors/sb/CS-023143.htm#l.
[140]维基百科编者.千禧年大奖难题维基百科,2010,http://zh.wikipedia.org/w/index.php?title=%E5%8D%83%E7%A6%A7%E5%B9%B4%E5%A 4%A7%E7%8D%8E%E9%9B%A3%E9%A1%8C&oldid=12750030.
[141]Malkawi A I H, Hassan W F, Sarma S K. An efficient search method for finding the critical circular slip surface using the Monte Carlo technique[J]. Canadian Geotechnical Journal. 2001(8):1081-1089.
[142]Cheng Y M. Location of critical failure surface and some further studies on slope stability analysis[J]. Computers and Geotechnics.2003,30(3):255-267.
[143]曾建潮,介婧,崔志华.微粒群算法[M].科学出版社,2004.
[144]Kennedy J, Eberhar R. Particle swarm optimization[Z]. Piscataway, USA:19951942-1948.
[145]Shi Y, Eberhart R C.A modified particle swarm optimizer[Z]. Piscataway NJ. IEEE Press, 1998.
[146]Clerc M. The swarm and the queen:towards a deterministic and adaptive particle swarm optimization.[Z]. Washington, DC:Piscataway, NJ:IEEE Service Center,19991951-1957.
[147]Shi Y, Eberhart R C. Empirical study of particle swarm optimization[Z]. Piscataway,NJ, IEEE Service Center,1945-1950.
[148]Chapman S. Fortran 95/2003 for scientists & engineers[M]. Columbus, OH., USA: McGraw-Hill,2007.
[149]Adams J C, Brainerd W S, Hendrickson R A, et al. The Fortran 2003 handbook:the complete syntax, features and procedures[M]. Minneapolis, Minnesota, USA:Springer,2008.
[150]Stepanov A, Mcjones P. Elements of programming[M]. Boston, MA., USA:Addison-Wesley Professional,2009.
[151]W L K S W. Rapid evaluation of the critical slip surface in slope stability problems[J]. International Journal for Numerical and Analytical Methods in Geomechanics.1987,11(5): 449-473.
[152]李守巨.基于计算智能的岩土力学模型参数反演方法及其工程应用[J].岩石力学与工程学报.2005(02).
[153]高玮,冯夏庭.基于免疫连续蚁群算法的岩土工程反分析研究[J].岩石力学与工程学报.2005(23).
[154]田明俊,周晶.基于蚁群算法的土石坝土体参数反演[J].岩石力学与工程学报.2005(08).
[155]Dorigo M, Stutzle T. Ant Colony Optimization[M]. Cambridge, MA, USA:The MIT Press, 2004.
[156]陈昌富.仿生算法及其在边坡和基坑工程中的应用[J].岩石力学与工程学报.2003(09).
[157]陈富昌,龚晓南,王饴孙.自适应蚁群算法及其在边坡工程中的应用[J].浙江大学学报(工学版).2003,37(5).
[158]Chapman B, Jost G, Kuck D J, et al. Using OpenMP:Portable Shared Memory Parallel Programming[M]. Cambridge, MA, USA:The MIT Press,2007.
[159]Seismosoft. SeismoSignal-A computer program for signal processing of strong-motion data[DB/CD].2004.
[160]南京工学院数学教研组.积分变换[M].北京：高等教育出版社,1989.
[161]Hayes M. Schaum's outline of digital signal processing[M]. Columbus, OH, USA: McGraw-Hill,1998.
[162]Stoica P, Moses R. Signals, Spectral Analysis Of Signals[M]. Pearson Education.
[163]Lyons R G. Understanding digital signal processing (2nd Edition)[M]. Upper Saddle River, New Jersey, USA:Prentice Hall,2004.
[164]陈育民,徐鼎平FLAC/FLAC3D基础与工程实例[M].北京：中国水利水电出版社,2009.
[165]Lysmer J, Kuhlemeyer L. Finite dynamic model for infinite media[J]. Engineering Mech.1969, 95(EM4):859-877.
[166]Kolsky H. Stress waves in solids[M]. New York:Dover Publications,1963.
[167]Itasca. FLAC3D Fast Lagrangian Analysis of Continua in 3 Dimensions User's Manual[M]. Minneapolis, Minnesota USA:Itasca Consulting Group Inc..
[168]Kuhlemeyer R L, Lysmer J. Finite Element Method Accuracy for Wave Propagation Problems[J]. Soil Mech.& Foundations, Div. ASCE.1973,99(SM5):421-427.
[169]Herlihy M, Shavit N. The Art of Multiprocessor Programming[M]. Burlington, MA, USA: Morgan Kaufmann,2008.
[170]Sun Microsystems Inc. Sun Studio 12:Fortran programming guide[G]. Santa Clara, CA, USA: 2007.
[2]殷跃平.汶川八级地震地质灾害研究[J].工程地质学报.2008(04)：433-444.
[3]殷跃平.汶川八级地震滑坡高速远程特征分析[J].工程地质学报.2009(02)：153-166.
[4]殷跃平.汶川八级地震滑坡特征分析[J].工程地质学报.2009(01)：29-38.
[5]Jibson R W, Harp E L, Michael J A. A method for producing digital probabilistic seismic landslide hazard maps[J]. Engineering Geology.2000,58(3/4):271-289.
[6]Keefer D V. Statistical analysis of an earthquake-induced landslidedistribution-the 1989 Loma Prieta, California event[J]. Engineering Geology.2000,58(3/4):231-249.
[7]Mahdavifar M R, Solaymani S, Jafari M K. Landslides triggered by the Avaj, Iran earthquake of June 22,2002[J]. Engineering Geology.2006,86(2/3):166-182.
[8]Wang W N, Nakamura H, Tsuchiya S, et al. Distributions of landslides triggered by the Chi-Chi earthquake in Central Taiwan on September 21,1999[J]. Journal of the Japan Landslide Society. 2002,38(4):18-26.
[9]晏同珍,杨顺安,方云.滑坡学[M].武汉：中国地质大学出版社,2000.
[10]Heim A. Bergsturz und Menschenleben[M]. Zutich:Naturforschenden Gesellschaft,1989.
[11]Scheidegger A E. On the prediction of the reach and velocity of catastrophic landslides [J]. Rock Mechanics and Rock Engineering.1973,5(4):231-236.
[12]Voight B. Rockslides and avalanches[M]. Amsterdam:Elsevier Scientific Pub. Co.,1978.
[13]Bock C G. Martinezmountain rock avalanche[C]. U.S.:Geological Society of America,1977. 155-168.
[14]Evans S G, Hungr O, Clague J J. Dynamics of the 19rock avalanche and associated distal debris flow onMountCayleBritish Columbia, Canada:Implications for landslide hazard sessment on dissected volcanoes[J]. Engineering Geology.2006,1(1):29-51.
[15]Qiao L, Yuan C, Miyajima M, et al. Shake-table testing and FLAC modeling of liquefaction-induced slope failure and damage to buried pipelines[C]. Sacramento, CA, United states:American Society of Civil Engineers,2008.
[16]Sun J, Tian X, Guan X, et al. Stability Analysis for Loosened Rock Slope of Jinyang Grand Buddha in Taiyuan, China[J]. Earth Science Frontiers.2008,15(4):227-238.
[17]He M C, Feng J L, Sun X M. Stability evaluation and optimal excavated design of rock slope at Antaibao open pit coal mine, China[J]. International Journal of Rock Mechanics and Mining Sciences.2008,45(3):289-302.
[18]Itasca. FLAC-3D fast Lagrangian analysis of continua in 3dimension version 3.0[R]. USA: Itasca ConsuItingGroup Inc.,2005.
[19]Huang X, Mao X, Miao X, et al. Study of slope reinforcement force based on FLAC[J]. Journal of China University of Mining and Technology.2006,16(2):123-127.
[20]Pan Q, Chang L. A study of application about FLAC-3D in damage of rock mass slope[J]. Key Engineering Materials.2006,306-308 II:1427-1432.
[21]崔芳鹏,胡瑞林,张明,等.重庆小南海地震崩滑体的基本特征及形成机制研究[J].工程地质学报.2009(02)：167-174.
[22]李维树,谭新,黄志鹏.长江小南海水利枢纽坝基弱风化岩体工程力学性质研究及可利用探讨[J].岩石力学与工程学报.2008(S2)：3899-3904.
[23]Kim D, Pan X, Pardalos P M. An Enhanced Dynamic Slope Scaling Procedure with Tabu Scheme for Fixed Charge Network Flow Problems[J]. Computational Economics.2006.
[24]Kim D, Pan X, Pardalos P. An Enhanced Dynamic Slope Scaling Procedure with Tabu Scheme for Fixed Charge Network Flow Problems[J]. Landslides.2006,27(2):273-293.
[25]刘红帅,薄景山,刘德东.岩土边坡地震稳定性分析研究评述[J].地震工程与工程振动.2005(01)：164-171.
[26]刘红帅,薄景山,刘德东.岩土边坡地震稳定性评价方法研究进展[J].防灾科技学院学报.2007(03)：20-27.
[27]祁生文,伍法权,刘春玲.岩质边坡动力反应分析[M].北京：科学出版社,2007.
[28]Lawphongpanich S. Dynamic Slope Scaling Procedure and Lagrangian Relaxation with Subproblem Approximation[J]. Doklady Earth Sciences.2006,35(1):121-130.
[29]Al-Homoud A S, Tahtamoni W W. Reliability analysis of three-dimensional dynamic slope stability and earthquake-induced permanent displacement[J]. Soil Dynamics and Earthquake Engineering.2000,19(2):91-114.
[30]Stewart J P, Blake T F, Hollingsworth R A. A Screen Analysis Procedure for Seismic Slope Stability[J]. Earthquake Spectra.2003,19(3):697-712.
[31]Lin M, Wang K. Seismic slope behavior in a large-scale shaking table model test[J]. Engineering Geology.2006,86(2-3):118-133.
[32]Urgeles R, Locat J, Lee H J, et al. The Saguenay Fjord, Quebec, Canada:integrating marine geotechnical and geophysical data for spatial seismic slope stability and hazard assessment[J]. Marine Geology.2002,185(3-4):319-340.
[33]Kokusho T, Ishizawa T, Nishida K. Travel distance of failed slopes during 2004 Chuetsu earthquake and its evaluation in terms of energy[J]. Soil Dynamics and Earthquake Engineering. 2009,29(7):1159-1169.
[34]唐辉明,滕伟福.岩土工程数值模拟新方法[J].地质科技情报.1998(S2)：42-49.
[35]De Santayana F P, Roque A J. Preliminary static and seismic slope stability analyses for a Portuguese municipal solid waste landfill[C]. Cardiff, Wales, United kingdom:Thomas Telford Services Ltd,2006.
[36]Urgeles R, Locat J, Lee H J, et al. The Saguenay Fjord, Quebec, Canada:Integrating marine geotechnical and geophysical data for spatial seismic slope stability and hazard assessment[J]. Marine Geology.2002,185(3-4):319-340.
[37]Bray J. Simplified Seismic Slope Displacement Procedures [J]. Geo-Marine Letters.2007: 327-353.
[38]Yegian M K, Harb J N. Slip displacements of geosynthetic systems underdynamic excitation[M]. Design and performance of solid waste landfills, No.54, Yegian M K, Finn W, San Diego, California, ASCE,:Geotechnical Special Publication,1995,212-236.
[39]Yegian M K, Lahlfa A M. Dynamic interface shear strength properties of geomembrabes and geotexiles[J]. J Geotech Engrg ASCE.1992,118(5):760-779.
[40]Terzaghi K. Mechanisms of landslides[C].1950.83-123.
[41]Seed H B. Considerations in the earthquake-resistant design of earth and rockfill dams[J]. Geotechnique.1979,29(3):215-263.
[42]Mononobe A H. Seismic stability of the earth dam[C]. Washinton DC:Ⅳ:1936.
[43]Seed H B, Martin R G. The seismic coefficient in earth dam design[J]. Journal of Soil mechanics and Foundation Division.1996,104(SM3):59-83.
[44]Newmark N M. Effect of earthquakes on dams and embankments [J]. Geotechnique.1965,15(2): 139-160.
[45]Wang K, Lin M. Development of shallow seismic landslide potential map based on Newmark's displacement:the case study of Chi-Chi earthquake, Taiwan[J]. Ecological Research.
[46]Ingles J, Darrozes J, Soula J. Effects of the vertical component of ground shaking on earthquake-induced landslide displacements using generalized Newmark analysis [J]. Engineering Geology.2006,86(2-3):134-147.
[47]Chang K, Taboada A, Lin M, et al. Analysis of landsliding by earthquake shaking using a block-on-slope thermo-mechanical model:Example of Jiufengershan landslide, central Taiwan[J]. Engineering Geology.2005,80(1-2):151-163.
[48]Lin J S, Whitman R V. Earthquake induced displacements of sliding blocks[J]. J Geotech Engng Div.1986,112(1):44-59.
[49]Makdisi F I, Seed H B. Simplified procedure for estimating dam and embankment earthquake induced deformations [J]. J Geotech Engng.1978,104(No.GT7):569-587.
[50]Martin R G, Finn W D L, Seed H B. Fundamentals of liquefaction under cyclic loading[J]. J Geotech Engrg Div asce.1975,101(GT5):423-438.
[51]Sarma S K. Seismic stability of earth dams and embankments[J]. Geotech Engng Div.1975,105, GT4:449-464.
[52]Sarma S K. Seismic displacement analysis of earth dams[J]. J Geotech Engng Div 107.1981, GT12:1735-1739.
[53]Wang S J, Xue S Y, Maugeri M. Dynamic stability of the left abutment in the Xiaolangdi Project On the Yellow River[C]. Istanbul(Turkey):A. a. Balma, Rotterdam,1993.
[54]薄景山,徐国栋,景立平.土边坡地震反应及其动力稳定性分析[J].地震工程与工程振动.2001(02)：116-120.
[55]陈玲玲,陈敏中,钱胜国.岩质陡高边坡地震动力稳定分析[J].长江科学院院报.2004(01)：33-35.
[56]Medina C L, Zeghal M. A micro-mechanical study of the dynamic response of unsaturated level sites and sloping grounds[C]. Denver, CO, United states:American Society of Civil Engineers, 2007.
[57]Lollino P, Santaloia F, Amorosi A, et al. Numerical analysis of the behaviour of a cut slope in stiff clays by finite difference method[C]. Graz, Austria:Taylor and Francis/Balkema,2006.
[58]Zou J Z, Willians D J, Xiong W L. Search for critical slip surfaces based on finite element method[J]. Canadian Geotechnical Journal.1995,32:233-246.
[59]Soares Jr. D. An optimised FEM-BEM time-domain iterative coupling algorithm for dynamic analyses[J]. Computers & Structures.2008,86(19-20):1839-1844.
[60]Wang X, Wang L B. Dynamic analysis of a water-soil-pore water coupling system[J]. Computers & Structures.,85(11-14):1020-1031.
[61]Assimaki D, Kausel E. Modified topographic amplification factors for a single-faced slope due to kinematic soil-structure interaction[J]. Journal of Geotechnical and Geoenvironmental Engineering.2007,133(11):1414-1431.
[62]Ford S A A S, Sitar N, Lysmer J, et al. Topographic effects on the seismic response of steep slopes[J]. Bulletin of the Seismological Society of America.1997,87(3):701-709.
[63]刘春玲,祁生文,童立强,等.利用FLAC-(3D)分析某边坡地震稳定性[J].岩石力学与工程学报.2004(16)：2730-2733.
[64]许瑾,郑书英.边界元法分析边坡动态稳定性[J].西北建筑工程学院学报(自然科学版).2000(04).
[65]孟达,孟祥冲,赵援.边坡的开挖及其地震动力响应分析[J].煤炭工程.2009(01)：77-80.
[66]杨扬.边坡稳定性分析方法简介[J].科技信息(科学教研).2008(03)：63.
[67]杨俊凯.边坡稳定性评价方法综述[J].中国科技信息.2009(16)：65-66.
[68]张云.露天矿边坡岩体裂隙含水介质渗流场边界元分析[J].南京大学学报数学半年刊.1992(01)：119-125.
[69]李建平,熊传治.岩石边坡安全系数的边界元解[J].岩石力学与工程学报.1990(03)：238-243.
[70]廖小平.一种新的边坡模拟技术[J].路基工程.1993(01)：43-47.
[71]樊文熙.有限元边界元法在岩体力学与工程中的应用[J].中州煤炭.1985(04)：2-3.
[72]Chunhan Z, Pekau O A, Feng J, et al. Application of disticnct element method in dynamic analysis of high rock slopes and blocky structures [J]. Soil Dynamics and Earthquake Engineering.1997,16(6):385-394.
[73]张建海,范景伟,何江达.用刚体弹簧元法求解边坡、坝基动力安全系数[J].岩石力学与工程学报.1999(04)：19-23.
[74]刘君,孔宪京,Shyu Kuokai. DDA与FEM耦合法在分缝重力坝非线性分析中的应用[J].计算力学学报.2004(05)：585-591.
[75]邓学晶,孔宪京,刘君.城市垃圾填埋场的地震响应及稳定性分析[J].岩土力学.2007,22(10)：2095-2100.
[76]刘君,孔宪京.卫生填埋场复合边坡地震稳定性和永久变形分析[J].岩土力学.2004,22(05)：778-782.
[77]卓家涛,章青.不连续介质力学问题的界面元法[M].北京：科学出版社,2004.
[78]裴觉民.数值流形方法与非连续变形分析[J].岩石力学与工程学报.1997(03)：80-93.
[79]Clough W R, Chopra K A. Earthquake stress analysis in earth dams[J]. J Engrg Mech, ASCE. 1966,92(2):197-211.
[80]谢康和,周建.岩土工程有限元分析理论与应用[M].北京：科学出版社,2000.
[81]李鸿波,张我华,陈云敏.爆炸冲击荷载作用下重力坝三维各向异性脆性动力损伤有限元分析[J].岩石力学与工程学报.2006(08).
[82]刘红帅,唐立强,薄景山,等.岩体弹粘塑性显式波动有限元分析[J].煤田地质与勘探.2009(01)：38-42.
[83]董志良,陈平山,莫海鸿,等.真空预压法有限元计算比较[J].岩石力学与工程学报.2008(11).
[84]熊敬,张建海Druker-Prager型屈服准则与强度储备安全系数的相关分析[J].岩土力学.2008,22(07)：1905-1910.
[85]郑宏,刘德富.弹塑性矩阵D_(ep)的特性和有限元边坡稳定性分析中的极限状态标准[J].岩石力学与工程学报.2005(07).
[86]许建聪,尚岳全,郑束宁,等.不平衡推力法的弹塑性有限元改进[J].岩石力学与工程学报.2005(23).
[87]王海斌,李永盛.边坡稳定性有限元分析的处理技巧[J].岩石力学与工程学报.2005(13).
[88]曾亚武,田伟明.边坡稳定性分析的有限元法与极限平衡法的结合[J].岩石力学与工程学报.2005(S2).
[89]孙世永.土动力学研究与发展综述[J].四川建筑.2006(03)：68-69.
[90]Cundall A P. A computer model for simulating progressive larg scale movement in blocky rock system[C]. Nancy France:1971.
[91]Itasca. Universal Distinct Element Code, Version 3.1 User's Manual[M]. Minneapolis, Minnesota USA:Itasca Consulting Group Inc.,1999.
[92]Bardet J, F S R. Seismic stability of fracture rock masses with the distinct element method[C]. Rapid City:1985.
[93]陶连金,苏生瑞,张倬元.节理岩体边坡的动力稳定性分析[J].工程地质学报.2001(01)：32-38.
[94]祁生文,伍法权,兰恒星.盘石头水库泄洪洞、导流洞进出口高边坡稳定性分析[J].岩石力学与工程学报.2002(03)：353-359.
[95]祁生文,伍法权,刘春玲,等.地震边坡稳定性的工程地质分析[J].岩石力学与工程学报. 2004(16)：2792-2797.
[96]祁生文,伍法权,孙进忠.边坡动力响应规律研究[J].中国科学E辑.2003(S1)：28-40.
[97]卢开澄,卢华明.组合数学[M].北京：清华大学出版社,2002.
[98]Xu Q, Fan X, Huang R, et al. Landslide dams triggered by the Wenchuan Earthquake, Sichuan Province, south west China[J]. Bulletin of Engineering Geology and the Environment.2009, 68(3):373-386.
[99]Tang C, Zhu J, Liang J. Emergency assessment of seismic landslide susceptibility:a case study of the 2008 Wenchuan earthquake affected area[J]. Acta Seismologica Sinica.2009,8(2): 207-217.
[100]Cui P, Zhu Y, Han Y, et al. The 12 May Wenchuan earthquake-induced landslide lakes: Distribution and preliminary risk evaluation[J]. Landslides.2009,6(3):209-223.
[101]王成华,夏绪勇.边坡稳定分析中的临界滑动面搜索方法评述[J].四川建筑科学研究院.2002,28(3).
[102]陈昌富,龚晓南.混沌扰动启发式蚁群算法及其在边坡非圆弧临界滑动面搜索中的应用[J].岩石力学与工程学报.2004(20)：3450-3453.
[103]Chen Y M. Location of critical failure surface and some further studies on slope stability analysis[J]. Computers and Geotechnics.2003,30(3):255-267.
[104]Zhu D Y. A method for locating critical slip surfaces in slope stability analysis[J]. Canadian Geotechnical Journal.2001,38:328-337.
[105]Mccombie P, Wilkinson P. The use of the simple genetic algorithm in finding the critical factor of safety in slope stability analysis[J]. Computers and Geotechnics.2002,29(8):699-714.
[106]Goh A T C. Genetic algorithm search for critical slip surface in multiple-wedge stability analysis[J]. Canadian Geotechnical Journal.1999,36:382-391.
[107]Boutrup E, Lovell C W. Search technique in slope stability analysis[J]. Engineering Geology. 1980,16(1):51-61.
[108]Siegel R A, Kovacs W D, Lovell C W. Random surface generation in stability analysis[J]. Journal of Geotechnical Engineering.1981,107(7):991-996.
[109]Baker R. Determination of the critical slip surface in slope stability computation[J]. International Journal for Numerical and Analytical Methods in Geomechanics.1980,4(4): 333-359.
[110]Pham H T V, Fredlund D G, Fredlund D G. The application of dynamic programming to slope stability analysis[J]. Canadian Geotechnical Journal.2003,40(830-847).
[111]曹文贵,颜荣贵.边坡非圆临界滑面确定之动态规划法研究[J].岩石力学与工程学报.1995(04)：320-328.
[112]陈云敏,魏新江,李育超.边坡非圆弧临界滑动面的粒子群优化算法[J].岩石力学与工程学报.2006(07)：1443-1449.
[113]李育超,凌道盛,陈云敏,等.蒙特卡洛法与有限元相结合分析边坡稳定性[J].岩石力学与工程学报.2005(11)：1933-1941.
[114]陈云敏,李育超,凌道盛.蒙特卡洛法与有限元结合搜索边坡临界滑动面[J].岩土力学.2004(S2)：75-80.
[115]Tingey B E, Mcbride J H,Thompson T J, et al. Study of a prehistoric landslide using seismic reflection methods integrated with geological data in the Wasatch Mountains, Utah, USA[J]. Engineering Geology.2007,95(1-2):1-29.
[116]Sutter H. The Free Lunch Is Over:A Fundamental Turn Toward Concurrency in Software[J]. Dr. Dobb's Journal.2005,30(3):16-20,22.
[117]张舒,褚艳利,赵开勇,等.GPU高性能运算之CUDA[M].北京：中国水利水电出版社,2009.
[118]维基百科编者.摩尔定律维基百科,2010,http://zh.wikipedia.org/w/index.php?title=%E6%91%A9%E5%B0%94%E5%AE%9A%E 5%BE%8B&oldid=12765148.
[119]Enet硅谷动力.英特尔80核心cpu研发成功2010,http://www.cz7v.com/news/zixun/201001/58.html.
[120]郑宏,田斌,刘德富,等.关于有限元边坡稳定性分析中安全系数的定义问题[J].岩石力学与工程学报.2005(13)：2225-2230.
[121]Zienkiewicz O C, Humpheson C, Lewis R W. Associated and non-associated visco-plasticity and plasticity in soil mechanics[J]. Geotechnique.1975,25(4):671-689.
[122]Lin H, Cao P, Li J, et al. Displacement characteristic of soil slope in shear strength reduction method based on Mohr-coulomb model[C]. China:Springer-Verlag GmbH and Co. KG,2008.
[123]Jiang Q. Strength reduction method for slope based on a ubiquitous-joint criterion and its application[J]. Mining Science and Technology (China).2009,19(4):452-456.
[124]刘金龙,栾茂田,赵少飞,等.关于强度折减有限元方法中边坡失稳判据的讨论[J].岩土力学.2005(08)：1345-1348.
[125]张鲁渝,郑颖人,赵尚毅.有限元强度折减系数法计算土坡稳定安全系数的精度研究[J].水利学报.2003(1)：21-26.
[126]张士兵,王练柱,张建.边坡稳定性弹塑性大变形有限元强度折减分析[J].岩石力学与工程学报.2004(S1)：4463-4467.
[127]赵尚毅,郑颖人,邓卫东.用有限元强度折减法进行节理岩质边坡稳定性分析[J].岩石力学与工程学报.2003(02)：254-260.
[128]郑颖人,赵尚毅,邓卫东.岩质边坡破坏机制有限元数值模拟分析[J].岩石力学与工程学报.2003(12)：1943-1952.
[129]杨有成.强度折减法在斜坡稳定性分析中的适用性研究[D].武汉：中国地质大学(武汉),2008.
[130]杨有成,李群,陈新泽,等.对强度折减法若干问题的讨论[J].岩土力学.2008,22(04)：1103-1106.
[131]陈新泽,唐辉明,杨有成,等.基于FLAC-(3D)强度折减法滑坡三维稳定性研究——以三峡库区白果树古滑坡群为例[J].水文地质工程地质.2008(02)：24-29.
[132]郑颖人,张玉芳,赵尚毅,等.有限元强度折减法在元磨高速公路高边坡工程中的应用[J].岩石力学与工程学报.2005(21).
[133]翟长海,谢礼立.结构抗震设计中的强度折减系数研究进展[J].哈尔滨工业大学学报.2007(08)：1177-1184.
[134]陈菲,邓建辉.岩坡稳定的三维强度折减法分析[J].岩石力学与工程学报.2006(12).
[135]段荣福,白世伟.基于临界滑动场的岩质边坡临界滑动面搜索方法[J].岩土力学.2007(05)：861-864.
[136]吴余生,陈胜宏.并行组合模拟退火算法在边坡稳定分析中的应用[J].岩土力学.2006(09)：1554-1558.
[137]李梦玄,余飞,王元汉.边坡失稳的仿真模拟分析[J].岩土力学.2002(S1)：169-171.
[138]陈祖煜.土质边坡稳定分析：原理·方法·程序[M].北京：中国水利水电出版社,2003.
[139]Intel.Intel微处理器出口规范测量机制,2010,http://www.intel.com/support/tw/processors/sb/CS-023143.htm#l.
[140]维基百科编者.千禧年大奖难题维基百科,2010,http://zh.wikipedia.org/w/index.php?title=%E5%8D%83%E7%A6%A7%E5%B9%B4%E5%A 4%A7%E7%8D%8E%E9%9B%A3%E9%A1%8C&oldid=12750030.
[141]Malkawi A I H, Hassan W F, Sarma S K. An efficient search method for finding the critical circular slip surface using the Monte Carlo technique[J]. Canadian Geotechnical Journal. 2001(8):1081-1089.
[142]Cheng Y M. Location of critical failure surface and some further studies on slope stability analysis[J]. Computers and Geotechnics.2003,30(3):255-267.
[143]曾建潮,介婧,崔志华.微粒群算法[M].科学出版社,2004.
[144]Kennedy J, Eberhar R. Particle swarm optimization[Z]. Piscataway, USA:19951942-1948.
[145]Shi Y, Eberhart R C.A modified particle swarm optimizer[Z]. Piscataway NJ. IEEE Press, 1998.
[146]Clerc M. The swarm and the queen:towards a deterministic and adaptive particle swarm optimization.[Z]. Washington, DC:Piscataway, NJ:IEEE Service Center,19991951-1957.
[147]Shi Y, Eberhart R C. Empirical study of particle swarm optimization[Z]. Piscataway,NJ, IEEE Service Center,1945-1950.
[148]Chapman S. Fortran 95/2003 for scientists & engineers[M]. Columbus, OH., USA: McGraw-Hill,2007.
[149]Adams J C, Brainerd W S, Hendrickson R A, et al. The Fortran 2003 handbook:the complete syntax, features and procedures[M]. Minneapolis, Minnesota, USA:Springer,2008.
[150]Stepanov A, Mcjones P. Elements of programming[M]. Boston, MA., USA:Addison-Wesley Professional,2009.
[151]W L K S W. Rapid evaluation of the critical slip surface in slope stability problems[J]. International Journal for Numerical and Analytical Methods in Geomechanics.1987,11(5): 449-473.
[152]李守巨.基于计算智能的岩土力学模型参数反演方法及其工程应用[J].岩石力学与工程学报.2005(02).
[153]高玮,冯夏庭.基于免疫连续蚁群算法的岩土工程反分析研究[J].岩石力学与工程学报.2005(23).
[154]田明俊,周晶.基于蚁群算法的土石坝土体参数反演[J].岩石力学与工程学报.2005(08).
[155]Dorigo M, Stutzle T. Ant Colony Optimization[M]. Cambridge, MA, USA:The MIT Press, 2004.
[156]陈昌富.仿生算法及其在边坡和基坑工程中的应用[J].岩石力学与工程学报.2003(09).
[157]陈富昌,龚晓南,王饴孙.自适应蚁群算法及其在边坡工程中的应用[J].浙江大学学报(工学版).2003,37(5).
[158]Chapman B, Jost G, Kuck D J, et al. Using OpenMP:Portable Shared Memory Parallel Programming[M]. Cambridge, MA, USA:The MIT Press,2007.
[159]Seismosoft. SeismoSignal-A computer program for signal processing of strong-motion data[DB/CD].2004.
[160]南京工学院数学教研组.积分变换[M].北京：高等教育出版社,1989.
[161]Hayes M. Schaum's outline of digital signal processing[M]. Columbus, OH, USA: McGraw-Hill,1998.
[162]Stoica P, Moses R. Signals, Spectral Analysis Of Signals[M]. Pearson Education.
[163]Lyons R G. Understanding digital signal processing (2nd Edition)[M]. Upper Saddle River, New Jersey, USA:Prentice Hall,2004.
[164]陈育民,徐鼎平FLAC/FLAC3D基础与工程实例[M].北京：中国水利水电出版社,2009.
[165]Lysmer J, Kuhlemeyer L. Finite dynamic model for infinite media[J]. Engineering Mech.1969, 95(EM4):859-877.
[166]Kolsky H. Stress waves in solids[M]. New York:Dover Publications,1963.
[167]Itasca. FLAC3D Fast Lagrangian Analysis of Continua in 3 Dimensions User's Manual[M]. Minneapolis, Minnesota USA:Itasca Consulting Group Inc..
[168]Kuhlemeyer R L, Lysmer J. Finite Element Method Accuracy for Wave Propagation Problems[J]. Soil Mech.& Foundations, Div. ASCE.1973,99(SM5):421-427.
[169]Herlihy M, Shavit N. The Art of Multiprocessor Programming[M]. Burlington, MA, USA: Morgan Kaufmann,2008.
[170]Sun Microsystems Inc. Sun Studio 12:Fortran programming guide[G]. Santa Clara, CA, USA: 2007.
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |