Computationally efficient stochastic optimization using multiple realizations

The presented study is concerned with computationally efficient methods for solving stochastic optimization problems involving multiple equally probable realizations of uncertain parameters. A new and straightforward technique is introduced that is based on dynamically ordering the stack of realizations during the search procedure. The rationale is that a small number of critical realizations govern the output of a reliability-based objective function. By utilizing a problem, which is typical to designing a water supply well field, several variants of this “stack ordering” approach are tested. The results are statistically assessed, in terms of optimality and nominal reliability. This study demonstrates that the simple ordering of a given number of 500 realizations while applying an evolutionary search algorithm can save about half of the model runs without compromising the optimization procedure. More advanced variants of stack ordering can, if properly configured, save up to more than 97% of the computational effort that would be required if the entire number of realizations were considered. The findings herein are promising for similar problems of water management and reliability-based design in general, and particularly for non-convex problems that require heuristic search techniques.     

基于快速非支配排序遗传算法的阻尼器多目标优化布置

将快速非支配排序遗传算法(NSGA-Ⅱ)和并行遗传算法相结合,提出内嵌NSGA-Ⅱ的粗粒度-主从式并行遗传算法。该算法将种群分为多个子种群,每个子种群可独立并行执行NSGA-Ⅱ操作;达到迁移周期时,子种群之间执行迁移操作;完成迁移后,子种群再次独立并行执行NSGA-Ⅱ操作。以最大层间位移角和最大楼层加速度为目标函数,对14层消能减震钢框架结构上的阻尼器布置位置进行优化分析。结果表明：该算法既实现多目标优化,又提高优化速度;对比常规隔层布置方法,该算法可使结构的层间位移角减震系数和加速度减震系数分别至少提高16.82%和16.01%。     

改进的遗传算法在地下水数值模拟中的应用

地下水流数值模拟中的模型识别问题，可以转化为函数的最优化问题。鉴于遗传算法的特点，将之引入到地下水流数值法中，用以解决地下水数学模型的识别问题。在建立地下水数值模拟中模型识别问题的是优化模型后，采取将最优化模型中的目标函数嵌入到遗传算法适应度函数中的方法，实现遗传算法与地下水流数值法的耦合。基于优化模型和遗传算法的运算过程，编写计算程序，实现地下水数学模型的自动识别。根据在珲春盆地地下水资源评价实例中应用得到的结果，信纸证了改进的遗传算法在地下水数值模拟中应用的可行性与有效性。     

A Review of: “Geographic Information Systems for Geoscientists—Modelling with GIS”

Distances from points to closest shorelines in a given direction are used, for example, in some models for estimating wave exposure. Such distances, also called fetch lengths, can be determined using standard geographic information systems. However, performance may be a problem if these distances are required for a great number of study points. Two new algorithms for determining fetch lengths for study points in the same directions are presented in this paper. It is assumed that the two‐dimensional map is stored in vector format, i.e. shorelines of islands and mainland are stored as polygons. The first algorithm works on a set of undirected line segments derived from the shoreline polygons. The other works on a raster representation of the map. The algorithm saves memory by postponing the rasterisation until necessary. Both of the new algorithms have superior efficiency to a previously reported algorithm when the number of study points is large.     

水库多目标供水调度规则研究

已有的水库供水调度规则研究偏重于单一目标供水,对于有着不同优先级和保证率的多个供水目标的研究不足.设计了一种水库多目标供水的调度规则,该规则能分别处理不同的供水目标,保证各项用水不发生深度破坏,在此基础上寻求累积缺水最小的调度目标.一种混合模拟和遗传算法的模型被用来对该规则的参数进行优化,在搜索规则最优参数组合的过程中遗传算法显示出很好的优化性能.实例研究检验了提出的供水调度规则的合理性.     

TSIP中期地震预报方法在辽宁地区的应用与研究

本文用ＣＮ算法研究了辽宁省两个地震重点监视防御区的中强震前的强震发生概率增长时间,结果表明预测效果较好,有震报准率为９０％,ＴＩＰ预测警戒占时空率为３０％左右,Ｒ值评分为０．６左右。在ＴＩＰ预测的基础上,进一步用ＳＩＰ方法分析这两个区域的空间背景发震概率的不均匀分布。综合应用ＴＩＰ方法和ＳＩＰ方法,对辽宁省及邻近地区进行了试验性地震预报。     

用遗传算法反演一维声波介质

局部线性化的反演方法不仅在初始模型远离真值时收敛速度慢,而且往往易陷入局部极大值中,而遗传算法(GA)便能解决这一问题,它是一种全局搜索法。虽然它同模拟退火法一样利用转移概率进行搜索,但其收敛速度一般比模拟退火法快。本文引进了多个目标函数进行综合评价和灾变过程,用单道地震道对一维声波介质的速度和密度同时进行了反演,并反演了加噪地震道和含薄层介质模型,取得了满意的效果。     

一种CMP道集剩余校正量求取算法

地震资料处理中CMP道集在经过静校正和动校正处理之后,仍然会存在剩余校正量。剩余校正量的存在影响叠加成像效果。传统剩余校正量求取采用基于统计规律的互相关算法。设计了一种振幅绝对误差剩余校正量求取算法,并用MATLAB编程进行了仿真测试和实际地震资料处理试验。两种算法有着相同的效果,但振幅绝对误差算法速度更快。     

基于阵列感应测井的支持向量机流体识别方法

基于研究区块的高分辨率阵列感应测井(HDIL)资料,首先分析了利用交会图法进行流体识别的效果,指出了该常规解释方法的局限性。在此基础上,将基于遗传算法优化的支持向量机(GA-SVM)引入到测井解释当中,结合阵列感应测井和常规测井资料,建立了流体识别的非线性模型。利用该模型对45个训练样本进行回判,准确率为100%,对21个预测样本进行预测的准确率为90.48%,相对于常规解释方法,该流体识别方法具有更高的准确率。