并行遗传算法对硅藻软岩本构关系的求解

利用并行遗传算法,基于不同条件下软岩三轴压缩实验的实验数据,对正常固结领域和超固结领域硅藻软岩的本构关系进行了拟合,建立了一种新的考虑时间效应的应力-应变-时间本构模型。预测结果分析表明,所建模型很好地反映了硅藻软岩的本构关系,从而验证了该模型并行遗传算法解的合理性和可行性。研究成果在地下空间的利用和开发中具有良好的应用前景,对介于岩石和土之间复杂软岩材料力学特性研究的进一步发展具有重要的参考价值。     

用十进制遗传算法识别泰斯模型参数

提出了用十进制遗传算法识别泰斯模型参数的方法和步骤,并给出了计算实例。本方法和传统的配线法与高斯牛顿法相比,具有更好的确定性和更高的精度。十进制遗传算法的主要优点在于采用十进制编码表达实际问题比较简单,算法的平均效率高,特别是对于高维和复杂优化问题。     

Parameter identification of minifrac numerical tests using a gradient boosting-based proxy model and genetic algorithm

In recent years, the petroleum industry has devoted considerable attention to studying fluid flow inside fracture channels due to the discovery of naturally fractured reservoirs. The behavior prediction of these reservoirs is a well-known challenging task, in which the initial stage consists of identifying reservoir hydromechanical parameters. This work proposes an artificial intelligence-based approach to identify hydromechanical parameters from borehole injection pressure curves acquired through minifrac tests. This approach combines proxy modeling with a stochastic optimization algorithm to match observed and predicted borehole pressure curves. Therefore, a gradient boosting-based proxy model is built to predict borehole pressure curves, considering a proper strategy to develop time series modeling. Moreover, a Bayesian optimization algorithm is applied to compute the gradient boosting hyperparameters. In this optimization scenario, this paper proposes an appropriate objective function established from the assumed time series prediction strategy and the k-fold cross-validation. Finally, a genetic algorithm is adopted to identify unknown hydromechanical parameters, solving an inverse problem. Based on the proposed workflow, a study of the importance of the hydromechanical parameters is developed. To assess the methodology applicability, the approach is employed to identify parameters in synthetic and field minifrac tests. The results present how this approach can adequately identify hydromechanical parameters of hydraulic fracturing problems.     

某砂土液化大变形本构模型参数的敏感性分析

基于Yang和Ahmed[1－5]等提出的砂土液化大变形本构模型,对该模型的硬化规则和弹塑性模量确定方法作了改进,把该本构模型扩展应用到三维液化大变形的数值分析中,实现了基于ABAQUS大型商用软件计算平台上砂土液化大变形的计算子程序开发。基于该计算平台,对该模型的主要参数在描述砂土液化动孔隙水压力增长和动应力-应变关系曲线等方面的可靠性和敏感性进行了研究。给出了模型全过程参数、剪胀过程参数、剪缩与剪胀状态转换点流动变形量控制参数对试样的应力-应变关系曲线的影响程度及其规律,并对模型的主要参数的敏感性进行了分析,所得结论为通过动三轴试验获得相关模型参数提供了有效的指导和帮助,同时也发展了砂土液化大变形新的数值计算方法。     

Soil parameter identification using a genetic algorithm

This paper is dedicated to the identification of constitutive parameters of the Mohr–Coulomb constitutive model from in situ geotechnical measurements. A pressuremeter curve and the horizontal displacements of a sheet pile wall retaining an excavation are successively used as measurements. Two kinds of optimization algorithms are used to minimize the error function, the first one based on a gradient method and the second one based on a genetic algorithm. The efficiency of each algorithm related to the error function topology is discussed. Finally, it is shown that the use of a genetic algorithm to identify the soil parameters seems particularly suitable when the topology of the error function is complex. Copyright © 2007 John Wiley & Sons, Ltd.     

土体渗流固结参数识别方法

根据土体固结过程中超孔隙水压力观测资料,建立了基于遗传算法的土体渗流固结参数非线性识别方法,解决了经典高斯-牛顿极小化问题所存在的局部极小问题和最小二乘法所存在的当初始值选择不合适时迭代过程发散的问题,提出了根据观测仪器的精度,建立迭代终止条件的方法。数值计算结果表明,本文所提出的非线性反演方法适合于土体固结参数识别等类似的反问题。     

蚁群算法与遗传算法融合及其在边坡临界滑动面搜索中的应用

临界滑动面搜索是边坡稳定性分析中一项非常重要的内容。相对于圆弧滑动面的确定,只需要圆心和半径3个未知量,非圆弧滑面的确定则需要找出若干个控制点,是一个多维空间的优化问题。非圆弧滑动面优化搜索问题相当复杂,常规优化算法往往达不到要求。改进了蚁群算法,使其具备在连续空间的搜索能力,并与遗传算法融合,形成优势互补,克服了遗传算法的无反馈能力导致无用的冗余迭代、求解效率低以及蚁群算法初期信息素匮乏导致算法速度慢的不足。通过与商用软件GEO-SLOPE的算例求解结果对比,来说明本算法的有效性。     

免疫遗传算法在土石坝筑坝粗粒料本构模型参数反演中的应用研究

建立了确定筑坝粗粒料本构模型参数的有限元计算理论与IGA相结合的反演分析方法,利用公伯峡筑坝现场大型载荷试验资料,对筑坝原级配料的本构模型参数进行了反演研究。结果表明,由于在微弱风化花岗岩掺加云母片岩,颗粒破碎严重,爆破堆石料（3BⅠ）的原级配粗粒料反演参数远低于室内三轴试验成果;与3BⅠ堆石料相反,砂砾料（3BⅡ）现场反演得到的邓肯E~B模型参数高于室内三轴试验值;室内三轴试验成果不能客观反映筑坝材料的力学特性,对于当今的300 m级高土石坝建设,宜加强对原级配粗粒料的力学特性研究;同时对原型实测的空间位移场不同拟合方式进行了探讨,认为仅考虑铅直向变形的IGA参数反演成果能基本满足土石坝工程的精度要求。     

CASM: a unified state parameter model for clay and sand

The purpose of this paper is to present a simple, unified critical state constitutive model for both clay and sand. The model, called CASM (Clay And Sand Model), is formulated in terms of the state parameter that is defined as the vertical distance between current state (v, p′) and the critical state line in v–ln p′ space. The paper first shows that the standard Cam-clay models (i.e. the original and modified Cam-clay models) can be reformulated in terms of the state parameter. Although the standard Cam-clay models prove to be successful in modelling normally consolidated clays, it is well known that they cannot predict many important features of the behavior of sands and overconsolidated clays. By adopting a general stress ratio-state parameter relation to describe the state boundary surface of soils, it is shown that a simple, unified constitutive model (CASM) can be developed for both clay and sand. It is also demonstrated that the standard Cam-clay yield surfaces can be either recovered or approximated as special cases of the yield locus assumed in CASM. The main feature of the proposed model is that a single set of yield and plastic potential functions has been used to model the behaviour of clay and sand under both drained and undrained loading conditions. In addition, it is shown that the behaviour of overconsolidated clays can also be satisfactorily modelled. Simplicity is a major advantage of the present state parameter model, as only two new material constants need to be introduced when compared with the standard Cam-clay models. © 1998 John Wiley & Sons, Ltd.     

基于改进遗传算法的边坡可靠度分析

采用基于概率的可靠性分析方法来分析边坡的稳定性,以克服目前大多数分析方法以确定性安全系数为评判标准的弊端。对于每次搜索中给定的滑移面,结合拉格朗日乘子法提出边坡可靠度的计算模型;采用遗传算法搜索关键滑移面,提出边坡的适应度函数。为克服标准遗传算法在进化过程中无法调节全局搜索和局部搜索速度、且可能陷入次优解的缺陷,改进了子代中最佳个体的确定方法,来并以相邻两代的群体多样性特征值增量为导向,动态调节交叉概率和变异概率,以保持群体的多样性和搜索的有效性。实例分析结果表明,本文方法可以提高遗传算法的搜索能力和收敛速度,保证解全局最优;该法适用于边坡稳定可靠度分析,且前处理简单;在遗传操作前预优化可靠指标计算模型能大大减少遗传算法中的计算量,节约机时。     

基于遗传算法的抗滑桩优化设计

抗滑桩由于其桩位布置灵活、施工比较方便、治理效果可靠,是滑坡治理工程中广泛采用的一种抗滑措施。一直以来抗滑桩设计都是沿用手工查表计算的方法,采用这种传统的抗滑桩设计方法不仅增加了工程师繁重的计算工作,而且很难实现优化设计。文章在对抗滑桩的工作机理进行分析的基础上,讨论了影响抗滑桩经济性的主要因素,以抗滑桩的截面宽、高和长度为基本要素,建立了抗滑桩优化设计的数学模型,并采用遗传算法进行求解。最后对一个工程实例用传统设计方法和优化设计方法所得的结果进行了比较,结果表明,对抗滑桩进行优化设计不仅能减轻工程设计人员的工作量,还可以在满足安全的条件下明显的降低工程费用,在滑坡治理工程中具有一定的实际意义。     

基于十进制遗传算法的含水层参数的反演

含水层参的反演是一个复杂的非线性优化问题,针对传统二进制遗传算法收敛性能差的缺陷,提出了反演含水层参数的十进制遗传算法。以直线隔水边界附近的井流模型为例,讨论了十进制遗传算法在含水层参数反演中的应用,并与二进制遗传算法的进行比较。结果表明,该方法在含水层参数的反演中不仅是可行的,而且具有较好的确定性和较高的精度;与二进制遗传算法相比,十进制遗传算法的收敛性较好,省时高效,且表示较为自然,容易引入相关领域知识。同时,结合实例的分析结果得出种群的规模对算法的收敛性没有明显的影响。     

剪胀性砂土本构模型的研究

对于中密砂和密砂,剪胀现象是一个比较显著的特征,相变线是描述剪胀现象的一条特征线,它比较容易确定。因此,以相变线作为状态参考线,提出并定义了一种状态参量,在已有弹塑性模型基础上,将状态参量引入剪胀方程和塑性硬化模量表达式中,建立了一个新的考虑砂土剪胀性的本构模型。该模型尤其适合于中密砂和密砂,对于松砂则退化为原弹塑性模型。模型形式较简单,包含9个材料参数,且比较容易确定。数值模拟结果与试验数据吻合较好。     

Development of hyperplasticity models for soil mechanics

Hyperplasticity theory was developed by Collins and Houlsby (Proc. Roy. Soc. Lon. A 1997; 453 :1975–2001) and Houlsby and Puzrin (Int. J. Plasticity 2000; 16 (9):1017–1047). Further research has extended the method to continuous hyperplasticity, in which smooth transitions between elastic and plastic behaviour can be modelled. This paper illustrates a development of a new constitutive model for soils using hyperplasticity theory. The research begins with a simple one‐dimensional elasticity model. This is extended in stages to an elasto‐plastic model with a continuous internal function. The research aims to develop a soil model, which addresses some of the shortcomings of the modified cam‐clay model, specifically the fact that it cannot model small strain stiffness, or the effects of immediate stress history. All expressions used are consistent with critical state soil mechanics terminology. Finally, a numerical implementation of the model using a rate‐dependent algorithm is described. Copyright © 2005 John Wiley & Sons, Ltd.     

Rational approach to anisotropy of sand

The paper presents a constitutive model for the three-dimensional deformation–strength behaviour of inherently anisotropic sand. Based on non-linear tensorial functions, the model is developed without recourse to the concepts in plasticity theory such as yield surface and plastic potential. Benefited from the fact that no decomposition of strain into elastic and plastic parts is assumed, a unified treatment of anisotropic behaviour of deformation and strength is achieved. Anisotropy is characterized by a vector normal to the bedding plane. The extension of the constitutive model is furnished by incorporating the vector under consideration of the principle of objectivity and the condition of material symmetry. Distinct features of the model are its elegant formulation and its simple structure involving few material parameters. Model performance and comparison with experiments show that the model is capable of capturing the salient behaviour of anisotropic sand. © 1998 John Wiley & Sons, Ltd.     

Bayesian model selection for sand with generalization ability evaluation

Current studies have focused on selecting constitutive models using optimization methods or selecting simple formulas or models using Bayesian methods. In contrast, this paper deals with the challenge to propose an effective Bayesian-based selection method for advanced soil models accounting for the soil uncertainty. Four representative critical state-based advanced sand models are chosen as database of constitutive model. Triaxial tests on Hostun sand are selected as training and testing data. The Bayesian method is enhanced based on transitional Markov chain Monte Carlo method, whereby the generalization ability for each model is simultaneously evaluated, for the model selection. The most plausible/suitable model in terms of predictive ability, generalization ability, and model complexity is selected using training data. The performance of the method is then validated by testing data. Finally, a series of drained triaxial tests on Karlsruhe sand is used for further evaluating the performance.     

A critical state sand plasticity model accounting for fabric evolution

Fabric and its evolution need to be fully considered for effective modeling of the anisotropic behavior of cohesionless granular sand. In this study, a three‐dimensional anisotropic model for granular material is proposed based on the anisotropic critical state theory recently proposed by Li & Dafalias [2012], in which the role of fabric evolution is highlighted. An explicit expression for the yield function is proposed in terms of the invariants and joint invariants of the normalized deviatoric stress ratio tensor and the deviatoric fabric tensor. A void‐based fabric tensor that characterizes the average void size and its orientation of a granular assembly is employed in the model. Upon plastic loading, the material fabric is assumed to evolve continuously with its principal direction tending steadily towards the loading direction. A fabric evolution law is proposed to describe this behavior. With these considerations, a non‐coaxial flow rule is naturally obtained. The model is shown to be capable of characterizing the complex anisotropic behavior of granular materials under monotonic loading conditions and meanwhile retains a relatively simple formulation for numerical implementation. The model predictions of typical behavior of both Toyoura sand and Fraser River sand compare well with experimental data. Copyright © 2013 John Wiley & Sons, Ltd.     

遗传算法在反演三维地下水流模型参数中的应用

本文以非均质各向同性承压三维非稳定流为理想模型,结合有限元法讨论了用遗传算法反演水文地质参数问题.计算结果表明,本文在简单遗传算法(SGA)的基础上提出的优体克隆+子体优生遗传算法(BCC-YGCD-GA)具有收敛速度快、解的精度高和避免出现早熟等优点.在水资源评价和矿床疏干计算中有广阔的应用前景.     

基于改进遗传算法的粘弹性岩体力学参数反演

把模式搜索嵌入目前广为应用的遗传算法中,使之和神经网络有机结合,提出了搜索—遗传—神经网络算法。该方法用经过最佳预测学习算法训练的神经网络来表达粘弹性岩体力学参数和位移之间的映射关系,除具有一般遗传算法的优点外,还提高了参数反演的精度,节省了参数反演的计算时间。结合某工程实例,验证了该方法在粘弹性岩体力学参数反演中的优越性。