Inverse Analysis of Deep Excavation Using Differential Evolution Algorithm

This paper presents the applications of the differential evolution (DE) algorithm in back analysis of soil parameters for deep excavation problems. A computer code, named Python‐based DE, is developed and incorporated into the commercial finite element software ABAQUS, with a parallel computing technique to run an FE analysis for all trail vectors of one generation in DE in multiple cores of a cluster, which dramatically reduces the computational time. A synthetic case and a well‐instrumented real case, that is, the Taipei National Enterprise Center (TNEC) project, are used to demonstrate the capability of the proposed back‐analysis procedure. Results show that multiple soil parameters are well identified by back analysis using a DE optimization algorithm for highly nonlinear problems. For the synthetic excavation case, the back‐analyzed parameters are basically identical to the input parameters that are used to generate synthetic response of wall deflection. For the TNEC case with a total of nine parameters to be back analyzed, the relative errors of wall deflection for the last three stages are 2.2, 1.1, and 1.0%, respectively. Robustness of the back‐estimated parameters is further illustrated by a forward prediction. The wall deflection in the subsequent stages can be satisfactorily predicted using the back‐analyzed soil parameters at early stages. Copyright © 2014 John Wiley & Sons, Ltd.     

图书情报资料工作在测绘教育中的地位与作用

学校图书馆是图书资料的供应与服务基地 ,其藏书量与藏书类别的多少 ,以及图书馆各专业情报资料源的数量与管理和服务水平 ,是办学条件的重要标志之一 ,因而 ,测绘图书情报资料的信息源建设、测绘科技文献的开发与利用和测绘科技情报资料为师生的服务 ,在测绘教育中有着举足轻重的地位与作用。     

A hybrid Bayesian back‐propagation neural network approach to multivariate modelling

There is growing interest in the use of back‐propagation neural networks to model non‐linear multivariate problems in geotehnical engineering. To overcome the shortcomings of the conventional back‐propagation neural network, such as overfitting, where the neural network learns the spurious details and noise in the training examples, a hybrid back‐propagation algorithm has been developed. The method utilizes the genetic algorithms search technique and the Bayesian neural network methodology. The genetic algorithms enhance the stochastic search to locate the global minima for the neural network model. The Bayesian inference procedures essentially provide better generalization and a statistical approach to deal with data uncertainty in comparison with the conventional back‐propagation. The uncertainty of data can be indicated using error bars. Two examples are presented to demonstrate the convergence and generalization capabilities of this hybrid algorithm. Copyright © 2003 John Wiley & Sons, Ltd.     

Marquardt算法磁测三维反演方法策略

在磁法三维反演中,反演对象数目往往巨大,受限于当今的计算机技术与计算方法,在超大规模方程组的求解时遇到极大的困难。所以,在解决实际问题时能力非常有限,对于Marquardt算法磁测三维反演问题尤为突出。为此,在这里提出了两个较为有效的方法策略,即将模型组体化和将组体分类,并在理论应用中得到了较好的效果。     

DCT域遥感影像的自适应遗传优化增强

本文在DCT域中提出了一种基于遗传算法的遥感影像增强技术,针对DCT域遥感影像增强参数需人工设定的问题,利用自适应遗传算法对其进行优化选取:设计了一种考虑偏差的适应度函数以驱动遗传算法进化,通过无分块思想的使用实现了无人工干预、无块效应的遥感影像增强。该方法不但保持了遥感影像的原始色调而且显著提升了遥感影像方差、平均梯度、信息熵等质量测度,有效改善了遥感影像的视觉效果,便于后续影像处理及应用。     

“雨燕”中风暴算法在北京奥运天气预报示范项目中的应用及改进

介绍了临近预报系统“雨燕”中的风暴系列算法, 包括风暴识别、 风暴追踪、 基于TREC技术的风暴位置预报, 以及预报位置实时评分算法等。利用2008年北京奥运期间出现的多个强对流天气, 统计出该风暴产品在30 min和60 min预报时效的绝对距离误差分别约为13 km和23 km。分析了北京奥运天气预报示范项目期间该风暴产品误差较大的原因, 主要集中于TREC技术本身及其适用范围, 以及风暴预报方案中处理细节的不足, 具体为TREC技术不适用于孤立的回波单体, 雷达探测边界对TREC技术的影响, TREC矢量有时呈现出一定的空间不连续性, 以及对孤立少动单体的不当处理等。针对上述原因, 提出了一系列的改进方案, 包括对用于风暴位置预报的TREC矢量增加一致性检验, 利用风暴的历史轨迹校正不恰当的TREC矢量; 对TREC技术中象素阵列大小进行统计分析, 选择最适合北京地区的阵列大小; 风暴预报位置超出回波范围时新的处理技巧等。利用北京奥运期间的强对流资料, 对改进后的风暴算法进行评估。结果表明, 一方面, 改进后的算法能较好地控制风暴预报位置的绝对误差, 在30 min和60 min预报时效, 绝对误差分别减小了25%和26%。另一方面, 由于预报位置精度的提高, 能够提升相邻时刻风暴匹配的效率, 使得与以前算法相比有更多的风暴样本参与了各个预报时效的评分。     

EXPERIMENTS TO TRACK STORMS USING MODERN OPTIMIZATION ALGORITHMS

Storm identification and tracking based on weather radar data are essential to nowcasting and severe weather warning. A new two-dimensional storm identification method simultaneously seeking in two directions is proposed, and identification results are used to discuss storm tracking algorithms. Three modern optimization algorithms (simulated annealing algorithm, genetic algorithm and ant colony algorithm) are tested to match storms in successive time intervals. Preliminary results indicate that the simulated annealing algorithm and ant colony algorithm are effective and have intuitionally adjustable parameters, whereas the genetic algorithm is unsatisfactorily constrained by the mode of genetic operations. Experiments provide not only the feasibility and characteristics of storm tracking with modern optimization algorithms, but also references for studies and applications in relevant fields.     

基于5分钟雷达资料降水估测模型的比较

For the Z-R relationship in radar-based rainfall estimation, the distribution of corresponding R values for a given Z value （or the corresponding Z value for a given R value） may be highly skewed. However, the traditional power-law model is physically deduced and fitted under the normal-distribution presumption of radar wave echoes associated with a rain rate value, and it may not be very appropriate. Considering this problem, the authors devised several generalized linear models with different forms and distribution presumptions to represent the Z-R relationship. Radar-reflectivity scans observed by a CINRAD/SC Doppler radar and 5-minute rainfall accumulation recorded by 10 ground gauges were used to fit these models. All data used in this study were collected during some large rainfalls of the period from 2005 to 2007. The radar and all gauges were installed in the catchment of the Yishu River, a branch of the Huaihe River in China. Three models based on normal distribution and a dBZ presumption of gamma distribution were fitted using maximum-likelihood techniques, which were resolved by genetic algorithms. Comparisons of estimated maximized likelihoods based on assumptions of gamma and normal distribution showed that all generalized linear models （GLMs） of presumed gamma distribution were better fitted than GLMs based on normal distribution. In a comparison of maximum-likelihood, the differences between these three models were small. Three error statistics were used to assess the agreement between radar estimated rainfall and gauge rainfall： relative bias （B）, root mean square error （RMSE）, and correlation coefficient （r）. The results showed that no one model was excellent in all criteria. On the whole, the GLM-based models gave smaller relative bias than the traditional power-law model. It is suggested that validations conducted in many previous works should have been made against a specific criterion but overlooked others.     

Integration schemes with substepping algorithms for creep analysis in geomaterials

This paper describes a robust and efficient methodology for predicting displacements, deformations, and stresses in geomaterials that are susceptible to creep. The methodology is based on two integration schemes, which consider substepping algorithms. The first scheme is used for integrating space-time relations in a global sense, whereas the second scheme is used for integrating stress-stain relations in a local sense. Different from previous studies, both integration schemes are easy to implement and general in the sense that they can be applied to any type of creep law. Through an in-house finite element simulator, several numerical tests are performed. They include triaxial and wellbore closure analyses considering soft soils and salt rocks. The results show that the combination of both schemes leads to stable and accurate solutions with reduced computational time.     

顶点Pk覆盖问题的研究综述

顶点覆盖问题是经典的NP完全问题,在排序、计算机网络等现实生活中有许多的应用.近几年来,许多研究者开始探究它的推广形式——顶点P_k覆盖（VCP_k）问题,即寻找一个顶点子集,从拓扑结构图中删除后使得剩下的顶点导出的子图不包含P_k路,其中P_k是指包含k个顶点的路.本文简单介绍了VCP_k问题的应用背景,归纳了它在近似算法、精确算法、参数化算法3个方面的主要研究进展,并分析了一些主要的方法和技巧.在此基础上,对VCP_k问题及其相关问题的研究前景进行了展望.