基于扩展模糊时间Petri网的故障诊断

针对故障诊断中可能出现的故障原因漏判问题,提出一种基于扩展模糊时间Petri网的故障诊断算法。该算法通过检查模糊时间戳确定故障时刻所有故障原因的状态,结合正向和逆向推理最大程度地解决故障原因漏判问题,从而确定主要、次要和其他故障原因。燃气轮机的故障诊断实例验证了该算法的有效性。     

基于知识模型的城轨信号故障诊断系统

目前,城市轨道交通(简称"城轨")信号维护支持系统仅能进行单一故障源报警,无法快速确定位故障原因和指导运维人员处置故障;而城轨信号系统故障种类繁多,诊断分析逻辑复杂,针对不同场景定制化开发故障诊断程序,无法快速响应运维需求且成本高昂.针对该问题,文章基于知识模型研发了一种信息化、平台化的城轨信号故障诊断系统,以实现信号...     

基于感知模糊Petri网的自动装弹机故障分析方法

坦克自动装弹机故障发生率高,原因复杂,为了缩短自动装弹机故障诊断时间,提高诊断准确度,在分析了自动装弹机内部原理及获取相关专家经验后,提出了一种基于感知模糊Petri网的故障诊断方法。结合自动装弹机的具体构造建立对应的PFPN故障模型,通过模糊Petri网表示故障传播过程,利用感知机误差反向传递方法学习有限的专家经验,确定Petri网络中触发事件的权值。通过正向推理,实现了对自动装弹机故障的准确判断。逆向推理结合最小割级方法,缩小排查范围,提高推理效率。以旋转故障机为例建立对应的 PFPN故障模型,与故障树模型诊断结果和历史诊断数据对比,验证该故障分析方法的合理性和有效性,实现了对自动装弹机的快速准确的故障诊断。     

直觉模糊Petri网的双向模糊故障推理算法

针对故障诊断中故障现象与故障原因之间复杂的不确定关系,利用直觉模糊集表达不确定性信息的优势和Petri网的图形处理问题的能力,给出了基于直觉模糊Petri网的双向模糊故障推理算法。该算法首先利用反向直觉模糊推理算法对模型进行约减,查找故障原因,再利用正向直觉模糊推理算法对模型进行计算,输出结果。该算法既可将故障信息化繁为简,降低推理过程的时间复杂度,还能够使故障诊断的确定性程度得到进一步提高。汽车发动机诊断案例表明了所给双向直觉模糊推理算法的可行性和有效性。     

卫星电源系统的有向图分层诊断方法

在卫星电源故障诊断问题的研究中,针对系统的知识获取难问题,提出了一种将有向图和CLIPS专家系统相结合的分层诊断方法。这种方法利用有向图知识描述容易和CLIPS推理引擎快速的优势,设计了一个以事实描述系统模型知识的自定义模板结构,同时采用有向图分层策略减小故障搜索空间,然后利用CLIPS提供的快速匹配算法完成诊断推理过程,最后结合已有的专家经验知识给出卫星电源的故障诊断结果。实验结果表明,新方法具有很好的准确性和快速性。     

论混合知识表示

所谓知识表示是为描述世界所作的一组约定,是知识的符号化。推理则是利用各类知识完成问题求解的过程。按推理中所使用知识的属性,推理可分为确定推理和不确定推理两类。确定推理是指推理过程中所使用的数据和知识是静态的、可靠的、确定的;不确定推理所使用的知     

基于故障知识库的机械零部件故障管理支持系统

 针对机械零部件故障知识表述困难,故障信息在不同设计人员间利用、分享和重 用效率低等问题,基于本体建模技术构建了故障知识库,设计开发了一个机械零部件故障管理 支持系统(FKMS)。通过机械零部件生命周期中不同角色对故障知识的需求分析,建立了系统的 功能结构模型,确定了故障知识的内容。利用本体建模技术构建了故障知识本体模型与实体关 系网络,保证了故障知识库结构的完备性。基于垂直模式建立了知识模型到数据库的映射,实 现了知识的高效存储与模型的可拓展。通过故障知识模型建立了故障原因关系网络,实现了故 障原因推理与置信度计算,解决了故障原因分析不全面的问题。基于 Django-Web 模块进行了 系统开发,并以直升机自由轮斜撑块涂层剥落故障为例验证了系统的知识管理和故障分析功能, 实现了对故障信息的分析与动态管理支持。     

基于机器学习的城市电梯困人故障原因预测方法研究

为实现电梯困人故障的应急处置快速响应,缩短现场故障原因排查时间,促进排障模式由人工经验向数据支撑下的智能诊断转变,利用梯度提升树算法（GBDT）建立电梯故障原因预测模型。经过数据清洗和特征提取,以2015—2020年南京市累积电梯故障数据进行模型训练。与真实值对比后的预测结果表明,前三位故障原因实时预测准确率可达81%...     

基于知识图谱的潜油电泵井故障诊断

潜油电泵井系统是油田开采重要工具,具有排量大、扬程高与作业环境灵活多变等优点.为了降低潜油电泵井系统故障危害,需要对其发生故障部件进行快速精确定位并维修.本文提出一种基于知识图谱的潜油电泵井故障诊断方法.采用改进BiLSTM-CRF实体识别算法与BERT关系抽取算法提取故障数据中的专家知识,构建潜油电泵井故障诊断领域知识图谱;利用构建知识图谱搭建以故障征兆为初始节点的贝叶斯推理网络,利用历史故障数据与条件概率解耦的计算方式推理出故障原因.本文通过故障诊断真实案例进行方法验证.     

基于OWL本体和SWRL规则的导弹智能故障诊断研究

针对复杂航空装备诊断知识缺乏、诊断效率低和知识共享性差等问题,以某型红外弹为例,提出一种基于OWL本体和SWRL规则的导弹智能故障诊断方法。首先以导弹FMECA结果作为知识源,通过基于ATML语法的OWL逻辑描述语言建立导弹本体模型,完成故障模式和故障原因本体之间的映射;其次采用语义网络规则语言SWRL描述知识库规则,建立本体知识单元之间类、属性和实例的对应关系,最后通过Racer推理机对导弹知识库进行故障诊断推理,获取故障诊断优先级顺序。推理结果表明,该方法能够解决复杂航空装备专家诊断系统中的知识表示困难、缺乏自动语义推理、重用共享性差等问题,获得最优的故障诊断路径的同时减少了故障排查步骤,从而实现了故障原因的快速定位,提高了复杂航空装备专家诊断系统的诊断效率和可靠性。     

An expert system for setting time steps in dynamic finite element programs

An expert system, ETUDES—Expert Time integration control Using Deep and Surface Knowledge System, which addresses the determination of the timestep for time integration of linear structural dynamic equations is described. This time-step may also be applicable for a moderately nonlinear simulation of the same structure. The program also determines whether an explicit or implicit method is most efficient for the particular simulation. A production rule programming system written in OPS5 is used for the implementation of this prototype expert system. Issues relating to the expert system architecture for this application, such as knowledge representation and structure, as well as domain knowledge are discussed. The prototype is evaluated by measuring it's performance in various benchmark model problems.     

TCN-KT：个人基础与遗忘融合的时间卷积知识追踪模型

智慧教育的热门领域知识追踪（KT）被视为序列建模任务,其主要关注点和解决方式都集中在循环神经网络（RNN）上。但RNN通常会面临梯度消失或者梯度爆炸等问题,且训练时间和设备要求都过于严苛,针对以上问题,提出融合学习者个人先验基础和遗忘因素的时间卷积知识追踪模型（TCN-KT）。首先利用RNN模型计算得到学生个人先验基础,然后使用梯度稳定、内存占用率更低的时间卷积网络（TCN）预测学生下一题正误的初始概率,最后融合基于学生基础的遗忘因素得到最终结果。实验验证,TCN-KT预测性能最佳并减少了计算时间。     

Formalizing knowledge on international environmental regimes: A first step towards integrating political science in integrated assessments of global environmental change

International environmental regimes are considered key factors in dealing with global environmental change problems. It is important to understand if and how regimes are effective in tackling these problems, which requires knowledge on their potential impact on these problems as well as on their political feasibility. Integrated assessments of global environmental change, which are mainly bio-physical and technology-economic oriented, barely address knowledge on environmental regimes, due to problems in drawing general and policy relevant lessons on regime effectiveness and inherent difficulties in modelling human and social dimensions. This paper presents an innovative approach to formalize knowledge on the effectiveness of environmental regimes, so that scientists from both the political science and integrated assessment domain can understand it, discuss it and contribute to it. We constructed a conceptual framework for the systematic analysis of conditions that influence regime effectiveness and implemented it in a computer model using fuzzy logic methodology. We evaluated the fuzzy model in an ex post case study on four existing international environmental regimes. The model can be used as an aid in analysing the effectiveness of existing or future regimes, highlighting which determinants contribute to success or failure, and it enables systematic and meaningful comparisons between regimes and policy measures.We discovered that formalizing knowledge on environmental regimes in a framework and model enhanced its transparency and deductive power as it forced us to be explicit about our choices and assumptions. Developing and using the framework and model also revealed the lacunae in knowledge in environmental regime theory which may inform regime researchers to further structure and increase their knowledge. By making knowledge on environmental regimes explicit and understandable we have taken an important step towards a better integration of political science in integrated assessments. We believe, however, that this integration is still in its early days and requires further attention in the future.     

通过最大化实现数据流算法中的可变滑动窗口(英文)

数据流挖掘中很多算法是基于定长滑动窗口的,定长滑动窗口的缺点是很难设置窗口的大小,而且对数据流分布的不同类型不存在最优大小的窗口,因此算法的性能较差。提出了可变滑动窗口算法,通过高效维护一个静态的最大范化均值完成。该常量在全部时间窗口中被最大化因而使用变长窗口。其他算法可以用该方法重新描述。实验表明了范化均值的有效性。     

Experimental results from parallel backward-chained expert systems

There are many applications which may be done by an expert system in real time, if the system is capable of real time response. the first Lisp- and Prolog-based expert systems have typically been too slow for real time response. This has lead to an effort to use other languages, the development of fast pattern matching techniques, and other methods of improving the speed of expert systems. Another approach to developing faster expert systems is to make use of the emerging parallel processing computer technology. A further use for parallelism is to allow reasonable response time for large knowledge bases. the size of knowledge bases may become as large as 20,000 chunks of knowledge (and more) in the near future in medical and space applications. This article describes the use of parallel processing in the EMYCIN backward chained rule-based model. Performance on two examples of shared memory multiprocessors is presented and contrasted with earlier simulations.     

A hybrid approach of differential evolution and artificial bee colony for feature selection

“Dimensionality” is one of the major problems which affect the quality of learning process in most of the machine learning and data mining tasks. Having high dimensional datasets for training a classification model may lead to have “overfitting” of the learned model to the training data. Overfitting reduces generalization of the model, therefore causes poor classification accuracy for the new test instances. Another disadvantage of dimensionality of dataset is to have high CPU time requirement for learning and testing the model. Applying feature selection to the dataset before the learning process is essential to improve the performance of the classification task. In this study, a new hybrid method which combines artificial bee colony optimization technique with differential evolution algorithm is proposed for feature selection of classification tasks. The developed hybrid method is evaluated by using fifteen datasets from the UCI Repository which are commonly used in classification problems. To make a complete evaluation, the proposed hybrid feature selection method is compared with the artificial bee colony optimization, and differential evolution based feature selection methods, as well as with the three most popular feature selection techniques that are information gain, chi-square, and correlation feature selection. In addition to these, the performance of the proposed method is also compared with the studies in the literature which uses the same datasets. The experimental results of this study show that our developed hybrid method is able to select good features for classification tasks to improve run-time performance and accuracy of the classifier. The proposed hybrid method may also be applied to other search and optimization problems as its performance for feature selection is better than pure artificial bee colony optimization, and differential evolution.     

Customer''s time-variant purchase behavior and corresponding marketing strategies: an online retailer''s case

The traditional customer relationship management (CRM) studies are mainly focused on CRM in a specific point of time. The static CRM and derived knowledge of customer behavior could help marketers to redirect marketing resources for profit gain at the given point in time. However, as time goes on, the static knowledge becomes obsolete. Therefore, application of CRM to an online retailer should be done dynamically in time. Though the concept of buying-behavior-based CRM was advanced several decades ago, virtually little application of the dynamic CRM has been reported to date.

In this paper, we propose a dynamic CRM model utilizing data mining and a monitoring agent system to extract longitudinal knowledge from the customer data and to analyze customer behavior patterns over time for the retailer. Furthermore, we show that longitudinal CRM could be usefully applied to solve several managerial problems, which any retailer may face.