一种基于属性变化的增量约简算法

随着处理数据工具的发展,许多数据库的属性都是动态增加的。当属性动态增加时,静态的约简方法消耗时间巨大,效率不高,针对这个缺陷,文中提出了属性变化的增量约简方法,通过增量技术计算增加属性后的知识粒度,然后在新的知识粒度和原有约简的基础上,能够在较短时间内找到新决策表的约简,仿真结果表明所提算法是有效的。     

基于灰色知识约简的城市供水工程项目排序决策指标优化研究

针对基于决策表的属性约简方法存在的问题,将灰色关联聚类法和粗糙集知识约简法相融合,利用灰色聚类实现数据的离散化和解决因决策属性不确定所带来的等价关系难确定问题;结合F统计量对条件属性集进行分类,构建基于灰色知识约简的指标优化体系模型,对城市供水工程项目排序决策的指标体系进行优化.以Z市供水工程项目排序决策的指标体系优化为例介绍模型的具体应用,计算结果表明,应用该模型能够保证指标筛选前后决策意图的一致性,提高后续排序决策的工作效率.     

粗糙集理论在故障诊断中的应用研究

粗糙集理论,凭借其在处理模糊和不确定信息上的优越性,已经被广泛引用于复杂系统的故障诊断中。对粗糙集理论以及其属性约简过程进行详细的介绍,并通过典型故障实例进行验证,诊断结果表明:在保证分类结果不变的前提下,粗糙集可以查找出对故障分类起主要作用的特征,从而达到属性约简的目的,为粗糙集理论在故障诊断中的深入运用打下了基础。     

粗糙集在CBR故障诊断中的应用

本文在对粗糙集属性约简算法进行论述的基础上,主要对粗糙集在CBR故障诊断中的应用进行了分析与探讨。     

基于RS理论的建筑工程成本预测研究

基于Rs理论，利用Rosetta软件对决策表进行完备化、离散化、属性约简等处理，确定影响建筑工程成本的6个关键因素，根据决策分类规则进行了建筑工程成本预测，获得了简单、易于理解的分类规则，为建筑工程成本预测提供了依据。     

基于粗糙集电梯群控系统调度算法研究

要解决电梯群控系统调度优化问题,首先应分析当前电梯群控调度算法存在的问题。以粗糙集理论为依据,提出以分辩关系强度进行群控电梯调度算法,算法将各种外部状态作为条件属性集、调度策略作为决策属性集形成信息决策表。对该表预先进行属性值约简后提得规则集,控制机将当前实际状态与规则集比较得到电梯群控指令。该算法突破了当前以经验知识推理、模糊逻辑、概率分析、仿生算法等调度算法的局限性,从外部状态和调度时机、调度目标给出该状态的控制规则,充分利用现有信息达到优化调度的目的,理论和实例相结合,证明了粗糙集理论解决电梯调度的合理性和有效性。     

基于改进PCA约简算法的研究

疾病的影响因素有很多,找到关键的影响因子对于前期疾病预测至关重要。文章数据来源是克利夫兰基金诊疗所关于心脏病诊断的数据,将样本数据与知识经验的方法相结合,利用改进的PCA对诊断数据进行属性约简并计算主成分权重,最终提取出关键的影响因子。     

基于粗糙集理论的属性约简算法研究

属性约简是粗糙集理论研究中的核心内容之一,现已证明寻找最小约简是NP-hard问题.在完备信息系统中,通过引入区分度的概念,以相对区分度为启发式信患,提出一种基于区分度的属性约简启发式算法,并对该算法的时间复杂性进行了分析.     

基于专家系统的消防控制故障诊断研究

本文分析了消防控制故障诊断的特性,运用专家系统提出了消防控制故障诊断的产生式规则知识获取和知识表示方法,设计出了消防控制故障诊断专家系统的模型。利用该系统模型已开发出了消防控制故障诊断专家系统原型。     

基于粗糙集和条件信息熵的高校教师绩效评价研究

针对当前高校发展的基础与需求,从影响高校教师评价结果的因素考虑,构建一个合理、高效的高校教师评价指标体系。将粗糙集理论和条件信息熵引入高校教师评价,建立基于粗糙集条件信息熵的综合评价智能模型,最大程度呈现专家的经验、知识对指标重要性的倾向。针对该体系中获取指标权重的问题,结合实际评价分析过程中小样本数据所呈现的特点,对决策表进行合理分块,从信息熵的角度结合粗糙集理论,以层次式计算的方式客观获取其权重值。通过实例分析,验证了该评价指标体系的合理性和粗糙集智能评价模型的有效性与优越性。     

Debugging Finite Element Programs Input Data with Machine Learning

Abstract: Debugging of the input data of a structural analysis program is a troublesome task which is heavily dependent on empirical knowledge. The paper describes an effort that applies machine learning to build an expert system for debugging faults in structural analysis program input data. ID3 decision tree induction algorithm is employed to build the automatic learning mechanism. This method Oflkrs an attractive potential for knowledge acquisition in the civil engineering domain in which expertise plays a dominant role. This paper is novel in at least three aspects: (1) machine learning is employed to build expert systems automatically; (2) a novel impurity function is proposed for splitting the decision tree; (3) a formulated comparison is proposed for evaluating learning results.     

Lightning protection decision system (an expert system)

Expert systems are fast becoming the leading branch of artificial intelligence technology. An expert system tool, ‘The Deciding Factor’, was investigated for use in building a Lightning protection decision system, an expert system for structures. The system described in this paper is a diagnostic expert system, developed for IBM PCs. There are a number of factors affecting the risk of the structure being struck by lightning and consequential effects of a stroke, which are capable of assessment. National Building Code of India has specified some judgemental and heuristic rules to combine these factors. The system was developed to model this expertise and to provide consultation to the user. This paper addresses the issue of knowledge engineering and demonstrates the ease with which the ‘knowledge engineer’ himself can build the systems with the use of an expert system shell.     

基于Geodatabase的长沙市基础空间数据库系统设计与实现

城市基础空间数据库系统是实现城市信息化建设和可视化管理的重要基础平台。本文首先介绍了Geodatabase数据模型,然后,以长沙市为例,分析了城市级空间数据系统的数据特点,阐述了“长沙市基础空间数据库系统”的数据库设计,包括数据分层设计、属性结构设计、数据编码设计等,最后,介绍了“长沙市基础空间数据库系统”的实现过程,并对系统的应用前景做了总结和展望。     

Acquisition, Representation, and Validation of Subjective Knowledge

Abstract: This paper concentrates on the three main stages of expert system development: knowledge acquisition, knowledge representation (which includes the treatment of uncertainties), and verification and validation. The findings of a survey of expert system applications in the construction management domain are reported. A model is presented that regulates the process of creating an expert system and that eliminates the shortcomings of existing techniques used with knowledge of subjective nature. This model is based on the Delphi method, a tool used for collecting data and information to solve nonanalytical problems. It is proposed that the use of the Delphi method in the knowledge acquisition, knowledge representation, and validation stages of expert system development will improve user confidence in the final product.     

Automation of Knowledge Organization and Acquisition

In this paper, an ongoing investigation on the development of SPERIL Systems, which are knowledge-based expert systems for damage assessment of existing structures, is described. The development of a new version, SPERIL-3, is presented and the motivations for the development of this system are discussed. Specifically, knowledge organization and knowledge acquisition in this system are emphasized. The automation of knowledge acquisition is presented and discussed.
In this paper, an ongoing investigation on the development of SPERIL systems, which are knowledge-based expert systems for damage assessment of existing structures, is described. The development of a new version, SPERIL-3, is presented and the motivations for the development of this system are discussed. Specifically, knowledge organization and knowledge acquisition in this system are emphasized. The automation of knowledge acquisition is presented and discussed.     

商业建筑初期灭火系统分析与决策

初期灭火是遏制商业建筑火灾的最佳时机,针对商业建筑火灾初期灭火过程中出现的问题,构建系统模型进行分析。数据来源于我国商业建筑特大火灾案例,采用贝叶斯统计分析方法进行分析,克服了火灾数据缺乏、且难以进行大量重复实验的困难,得到参数的条件期望估计与方差。通过对系统故障树的分析,得到我国商业建筑火灾初期灭火失败概率的期望与方差,并对参数引起的误差进行系统误差传播分析,最后进行系统决策。     

用CLIPS实现基于二叉树知识的汽车故障诊断

故障树是一种倒置的逻辑关系图 ,具有标准的知识结构 .同时也是故障诊断专家系统问题求解策略的形式化方法 .基于这种形式化方法的故障诊断专家系统可在一定程度上减小知识表示和获取的难度 .主要介绍了如何用 CLIPS语言实现基于二叉故障树知识的汽车故障诊断推理     

面向故障诊断的供水管网水压监测点优化布置方法

为解决供水管网故障诊断中准确率不高、经济性不佳的问题,设计了一个基于改进的果蝇优化算法核极限学习机的供水管网故障诊断模型。经验证,该模型具有学习速度快、故障识别率高等优点。以该模型为核心提出了基于果蝇优化算法的供水管网水压监测点优化布置方法。该方法首先利用果蝇优化算法形成多组水压监测点方案,然后用供水管网故障诊断模型计算每种方案的诊断准确率,选择其中诊断准确率最高、经济性最好的方案作为候选最优方案,并以此方案为基础,使用果蝇优化算法不断循环迭代,最终找到故障诊断准确率高且经济性最好的水压监测点布置方案。利用Matlab语言代码对提出供水管网水压监测点优化布置方法进行了编码实现,使用实际管网数据进行了实验,结果表明,所提出的供水管网水压监测点优化布置方法是一个有效的水压监测点优化布置方法。     

Ultrasonic imaging of microscale processes in quartz gouge during compression and shearing

Active ultrasonic monitoring in rock joints and gouge materials has the potential to detect the signatures of shear failure for a wide range of sliding modes, from slow and stable movements to fast and unstable sliding. While these collected measurements currently are being used to identify the seismic precursors to shear failure in rock joints and gouge materials, the underlying physical processes and contact scale mechanisms that control the changes in ultrasonic wave attributes are still poorly understood. To address this knowledge gap, this paper aims to investigate the relationship between the variations in ultrasonic wave attributes and the underlying particle scale mechanisms during both compression and shearing. Our double direct shear experiments were coupled with ultrasonic wave propagation measurements on granular quartz gouges, in which the gouge layers were sheared under different sliding velocities and constant normal stress conditions. Simultaneously, ultrasonic waveforms were continuously recorded during shearing with a fast data acquisition system and three pairs of ultrasonic wave transducers embedded at the two sides of the gouge layers. Different particle comminution mechanisms were observed from the non-uniform distribution of normal and shear stresses through the changes in ultrasonic transmissivity and scanning electron microscope (SEM) images. Our results show that the signatures of the geometry- and time-dependent variations of the inter-particle contact quality and pore volume changes with sliding velocity and slip accumulation were clearly captured from the variations in the transmitted wave amplitude and the dominant frequency, respectively. In addition, we found that variations in dominant frequency corresponded to dilation and compaction of the granular gouge layer during compression as well as stable and unstable sliding. Our results therefore confirmed that non-destructive acoustic techniques are capable of capturing a variety of micromechanical processes during fault gouge deformation and may prove useful in natural settings.     

Assessing reliability of design,construction, and safety related decisions

Currently there is no approach which would help to comprehensively ensure occupational safety. Many scientists perform researches and calculations, create new methods related to safety and health, but most of them analyze separate aspects of safety in the field of construction. The authors of this paper present a new complex view on ensuring occupational safety and health during construction. The selection of safety solutions is performed based on complex evaluation of structure, technology and safety. In their previous works, the authors offered a new method for multiple attribute decision synthesis, SyMAD-3, which helps to choose an effective construction project alternative from multiple alternatives by assessing various construction, technological and occupational safety solutions, based on a set of quantitative attributes. However, the integration of these solutions may cause doubts, since decision making in construction is always associated with uncertainty. The investment projects in construction are characterized by the large accuracy variation (from 15 to 50%) of some attribute values. Although the SyMAD-3 method is mathematically grounded, it does not answer the question if the error of attribute values impacts the final decision and if this decision can be reliably assessed.In the present paper, the authors supplement the SyMAD-3 method with decision sensitivity analysis (SyMAD-3 with SA) to improve the reliability of the SyMAD-3 method and assess the reliability of the obtained decision. The SyMAD-3 with SA method allows us to choose an effective alternative of a construction project by assessing three stages of construction, based on a set of attributes given the error of their values, and determine the reliability of the final decision. The proposed method is implemented in a software package created by the authors with the aim of analyzing decisions and performing experimental calculations in the field of construction.