人工神经网络在基桩低应变完整性检测中的应用

目前基桩低应变完整性检测数据的后期处理有很多方法 ,但分析中人为干预较多。利用人工神经网络强大的非线性映射能力和学习训练功能 ,提出了基于BP网络的基桩完整性检测模型。该模型基于现场实测资料 ,避免了数据处理过程中各种人为干预。应用该模型对工程实例进行了分析 ,训练和测试网络结果说明该方法能够快速、方便地对基桩质量进行模式识别     

一次逼近理论下边坡稳定的模糊随机数值分析

在协调元基础上,通过一次逼近理论,建立随机有限元模型,并通过模糊数学模型实现对边坡体工作状态的软化处理;借助于模糊随机数值方法对荆南长江干堤局部破坏失效进行了较为全面的研究。     

水资源承载能力的模糊综合评价研究

水资源承载力涉及到整个资源、经济、环境大系统,其是否协调是全球关注的重大问题.充分理解水资源承载力的概念,根据水资源承载力分析的结构特点,在传统多目标分析决策技术的基础上,分析了承载力评价的多级模糊综合评判方法,建立了水环境承载力多级模糊综合评判评价模型,并应用于四川省都江堰灌区水资源承载力评价.结果表明,提出的多级模糊综合评判由于综合考虑了评价因素的层次性和模糊性,其结果较一级模糊综合评判方法更符合实际.     

Prediction of Sandstone Porosity Through Quantitative Estimation of its Mechanical Compaction During Lithogenesis (Bibieybat Field, South Caspian Basin)

In this paper a new and easy quantitative approach based on an exponential decrease in intergranular volume as a function of effective stress and incorporated with fuzzy mathematics is suggested for evaluation of the lower limit of sandstones porosity. Furthermore, the comparison of predicted values of sandstones porosity with factual ones allow drawing some conclusions regarding succession of compaction and cementation processes taking place through burial history of rocks.     

双目标耦合不确定性模糊模式识别模型

本文在陈守煜建立的模糊模式识别理论的构架基础上,将加权广义距离与模糊熵最小作为模式识别的复合目标函数,建立了考虑随机和模糊不确定性,使目标函数最小的新型模糊模式识别耦合模型。应用此模型对我国12个湖库的富营养化程度进行综合评价,并与原模型的双目标结果进行比较,说明本文建立的模型考虑了不确定性因素的优越性。     

Fuzzy topological relations between fuzzy spatial objects

Fuzziness is an internal property of spatial objects.How to model fuzziness of a spatial object is a main task of next generation GIS.This paper proposes basic fuzzy spatial object types based on fuzzy topology.These object types are the natural extension of current nonfuzzy spatial object types.A fuzzy cell complex structure is defined for modeling fuzzy regions,lines and points.Furthermore,fuzzy topological relations between these fuzzy spatial objects are formalized based on the 9intersection approach.This model can be implemented for GIS applications due to its scientific theory basis.     

副热带高压强度变化的模糊聚类诊断预测

采用模糊C均值聚类 (FCM)、遗传算法 (GA) 和模糊减法聚类 (FSC) 交叉融合、优势互补思想进行副热带高压影响因子的综合聚类分析和副热带高压指数的诊断预测。在统计分析的基础上, 通过选择若干与副热带高压指数关系密切的影响因子构成了高维特征空间, 进行了综合聚类分析, 实现了副热带高压指数的聚类判别和诊断预测。该文提出的综合聚类方法既可克服FCM/GA算法全局/局部寻优的不足, 又可客观确定聚类数目。试验结果表明, 该方法具有良好的分类效果, 判别结果与实况基本相符。     

Analysis of sinkhole occurrences over abandoned mines using fuzzy reasoning: a case study

In Korea most of the old mine workings were worked with room and pillar method or sublevel caving method and today they possess great possibility of surface subsidence especially for shallow depth mines. In most of the cases, mine roadways, rooms and pillars are irregular in shape and information about the local geology is uncertain. For these reasons, it is difficult to standardize the estimating method of subsidence especially sinkhole type over abandoned mine area. This paper describes the application procedure for the fuzzy reasoning techniques to analyze the possibility of sinkhole occurrences over abandoned mines. This technique is implemented in software which can simplify the analysis procedure and present the possibility of sinkhole subsidence without having precise information about local geological/mining conditions. This technique has been applied to forecast sinkhole possibilities at Bonghwa site where a massive sinkhole has already been occurred.     

GIS中空间数据不确定性的混合熵模型研究

基于信息理论和模糊集合理论，针对GIS中部分空间数据既具有随机性又具有模糊性的特点，建立了空间数据不确定性的混合熵模型。以GIS中线元不确定性为例，讨论了线元不确定性的统计熵、模糊熵和混合熵估计方法，并针对特例给出了线元不确定性的熵带分布。     

基于模糊神经网络的土地合理储备量预测研究

将神经网络和模糊理论相结合建立模糊神经网络模型,从模糊神经网络角度并运用灰色系统理论对建设用地量进行预测,并应用于重庆市2005～2010年的建设用地量预测.计算分析结果表明,该模型具有良好的可行性和合理性,可以为确定土地合理储备量提供依据.