竖炉焙烧过程的智能设定模型

为了改善多目标评价案例推理设定模型在竖炉焙烧过程控制中的性能,运用注水原理分配过程变量的权重和群决策修正方法对多目标评价案例推理设定方法进行改进,得到一种新的智能设定模型.首先引入注水原理构造Lagrange函数对过程变量的权重进行优化分配,再通过案例检索和案例重用得到设定值的建议解,并根据多目标评价模型预测建议解对生产指标的影响效果,最后,对不合理的设定值进行群决策修正.将得到的设定模型应用于竖炉焙烧过程控制中,通过实验测试和对比应用说明了本文方法优于其他方法,能够有效提高多目标评价案例推理设定模型的控制性能.     

基于蒙古语课程领域语义Web的推理与检索方法的研究

由于蒙古语网络教学相对落后,蒙古语教学资源的重用和利用率较低,并且基于关键字的教学资源检索缺少在语义层面上的处理和表示,导致检索结果不全面和不准确。针对这种情况,提出基于蒙古语课程领域语义Web的检索方法。通过建立教学领域"人工智能(■■)"课程蒙文本体,利用关系数据库MySQL、语义相似度以及推理规则等工具与方法,实现构建基于蒙古语教学领域查询与检索原型系统。实验结果表明,该方法能有效地提高信息检索的查全率和查准率,并且建立了蒙古语教学资源可以共享和互操作的知识层次结构。     

基于案例分析金字塔模型的主动运维案例体系建设

针对传统运维案例在经验共享和处置效率方面的不足及运维领域同类事件重复发生概率高、处置差异大的情况,提出一种案例分析金字塔模型,使用基于向量化解析和知识分层提取的素材库构建方法,实现由生产运维操作标准库、生产运行异常事件库到案例素材库再到案例教学库的自动生成转变,并以金字塔模型为基础,通过引入改进的趋势预测分析方法、案例推理"5R"模型和"4S"案例教学模式完善预警预测、辅助决策、案例教学相关应用,建立主动运维案例体系。实践结果表明,通过整个体系的推行,改变了企业信息系统运维被动应付的局面,同类事件占比逐年降低,事件响应时间和处置时间明显缩短,运维服务质量大幅提升。     

Construction-specific spatial information reasoning in Building Information Models

In recent years, there have been significant advances in modeling technology for object-oriented building products. However, the building models are still lacking of providing construction-specific spatial information required for construction planning. Consequently, construction planners visually analyze building product models and derive geometric characteristics such as bounded spaces and exterior perimeter to develop detailed construction plans. Such a process presents fragmented information flows, from building product information to construction planning, that rely on subjective decisions of construction planners. In order to overcome these drawbacks, this research proposes a geometric reasoning system that analyzes geometric information in building designs, derives the construction-specific spatial information, and uses the information to assist in construction planning. The scope of presented work includes detecting work packages formed by faces during construction, such as large work faces and bounded spaces, and using information in the work packages directly to support planning of selected indoor construction activities. The main features of the proposed system named Construction Spatial Information Reasoner (CSIR) include a set of relationship acquisition algorithms, building component relationship data structure, and interpretation of the relationship to support detailed construction activity planning. The relationship acquisition algorithms identify adjacency between building components that is stored in the relational data structure. Then, acquired adjacency relationships are transformed into a set of graphs that represent work packages. To implement the proposed approach, CSIR utilized a commercially-available Building Information Modeling (BIM) platform and the algorithms were imbedded to the BIM platform. For validation, CSIR was tested on a real commercial building. For interior ceiling grid installation activities, CSIR successfully detected existing work packages and analyzed the spatial characteristics impacting construction productivity. The major contribution of the presented research would be to enable a realistic analysis of building geometric condition that is not possible in current BIM and a seamless information flow from building product information to construction process plans. These can potentially reduce current manual and error-prone construction planning processes. Limitations and future research suggestions are also presented.     

基于CBR的信息化装备故障案例推理复用技术研究

针对信息化装备技术保障要求,在对信息化装备故障案例特点进行分析的基础上,运用CBR理论解决信息化装备故障诊断的实际问题,提出了信息化装备故障案例推理复用策略,为构造实用的信息化装备故障诊断专家系统提供支持,对加强信息化装备技术保障具有积极的意义.     

基于粗糙集数据分析的可拓推理机制研究

目前可拓变换推理中的可拓变换主要依靠历史资料、人为指定或过往经验来进行,这大大制约了可拓变换在智能化推理中的应用。为解决此问题,提出了一种基于粗糙集数据分析的可拓推理机制。该机制首先采用粗糙集对数据进行分析来获取分类及规则知识,然后利用这些知识来指导可拓推理,从而实现了可拓变换及推理的可控性和高效性,为可拓推理的智能化应用莫定了基础。     

基于Value双向近似推理的系统决策方法

先提出Vague值间的一种新的相似度,并基于有序加权聚类(OWA)算子给出Vague集间的新的加权相似测度,其中的权重是由决策者根据实际决策需求来确定的,具有一定的弹性;另外通过Vague输入事实与Vague决策规则的前件(或后件)之间的相似匹配度,得到一个简单可行的双向近似推理方法,这可为复杂环境下的系统决策(如工业检测和医疗诊断等领域问题)提供有效的支持工具。     

系统L在非均匀概率空间下命题的真度理论

将经典二值命题逻辑L中公式的真度概念推广到势为2的非均匀概率空间上;当p∈(0,1)时,证明了全体公式的真度值之集在[0,1]中没有孤立点;利用真度定义公式间的p-相似度和伪距离,进而定义了p-逻辑度量空间,证明了该空间没有孤立点,并在此空间中提出了三种不同类型的近似推理模式。     

基于产生式探井决策专家系统的研究与应用①

基于产生式的探井决策专家系统以人工智能为理论,采用产生式知识表示方法,以油井钻探数据为依据,将获取的探井专家经验知识存入知识库,通过高效的推理,给出合理的决策方案,提高探井工程效率。