基于故障树分析的核动力装置实时智能故障诊断专家系统设计

为实时有效诊断核动力装置故障,提高装置可靠性,本文在传统专家系统的基础上引入故障树分析技术,设计了基于综合诊断重要度的推理机制,引入机理模型诊断技术对专家系统诊断结果进行验证,进一步提高诊断系统的可靠性。     

基于模糊推理的核供热堆故障诊断系统研究

介绍了基于模糊推理的核供热堆故障诊断系统的实现。工作结果表明,对于基于规则的推理系统来说,采用模糊推理的方法使推理结果较之传统的方法让反应堆运行人员更易接受。     

EAST低温系统的故障树分析

EAST是世界上进行聚变研究的先进超导托卡马克实验装置.低温冷却系统是EAST的主要子系统之一,担负着纵向和极向场线圈、纵场线圈盒、冷屏和电流引线的冷却功能,需保证其能长期稳定运行.本文运用故障树分析法来评价低温系统的可靠性,使用自主开发的软件RiskA建立低温系统的故障树模型并进行定量计算,得到系统的失效概率和部件的重要度,并提出改进建议,为系统的优化设计和提高可靠性提供参考.     

核测井系统故障诊断

为快速诊断和排除核测井系统应用现场出现的各种故障,提出了基于故障树、基于模糊理论和基于专家系统的故障诊断方法,给出的故障诊断实例表明,论文的故障诊断方法能够满足现场故障诊断和排除的需要,指出了测井系统故障诊断的发展方向.     

基于故障树的核电厂保护系统电路可靠性预计

可靠性预计作为电路可靠性分析的一种手段,在核电厂保护系统改造中,针对元器件停产、再选型具有很好的辅助作用。在电路可靠性分析中,可靠性预计主要是对每一种部件进行失效模式和影响分析(FMEA)评估,建立失效模型,计算失效率。故障树作为可靠性预计的一种分析方法,能直观体现卡件的单点故障。通过故障树法特有的树状结构图列出卡件的失效率计算式,从而得出电路改造前后卡件的平均无故障工作时间(MTBF)并将结果进行对比,该对比结果能作为核电厂保护系统电路改造可靠性分析的一项重要指标。     

核电站机组故障诊断系统的知识获取

核电站机组故障诊断系统知识的获取和知识库的建立是影响诊断系统能否应用于实际的关键步骤。针对实现核电站机组故障诊断系统给出了知识获取的一种方法和步骤,使其有章可循,加强了在实际核电站中可操作性。按照文中提出的工作框架组织人员完成各项任务,可以最终完成核电站机组故障诊断系统知识库的建立。     

HWRR控制区排风系统故障树分析

采用故障树分析法对HWRR控制区排风系统进行了分析,计算了排风系统的失效率,分析了影响排风系统失效率的主要因素。分析结果表明,交流母线和检修人员是影响排风系统可靠性的两大主要因素,应该予以重视。     

基于多层流模型和故障树的可靠性分析方法研究

多层流模型(MFM)和故障树以不同的形式描述系统知识,在相同的系统边界条件和假设下,两者表达的系统可靠性逻辑是等效的。本文工作以此为基础,结合MFM的特点,提出了MFM转换为故障树的方法,为快速建立故障树提供了一种途径,实现了基于MFM的可靠性定性分析,并以压水堆核电厂的安全注入系统为例建立了系统的MFM,定性地分析了系统的可靠性。分析结果表明,MFM转换为故障树的逻辑是正确的,且MFM易于理解、建立和修改,相对于传统建故障树的方法,大幅减少了分析人员的工作量,节省了建模时间。     

基于人工免疫系统的核动力设备故障诊断

针对核动力设备的复杂性和特殊性,提出了基于人工免疫系统的核动力设备故障诊断方法,给出了运用反面选择算法进行建模和克隆变异进化算法进行进一步故障识别的原则.将在旋转机械上采集到的某种信号进行特征提取,并将其作为信号的特征向量输入到建立好的AIS模型.仿真结果表明,此模型能较好地对各种故障类型进行识别.     

基于故障树分析的AP1000非能动余热排出系统的设计与优化

非能动余热排出(PRHR)系统是AP1000核电站重要安全系统之一。采用故障树方法对该系统可靠性进行评价,得到系统可能失效机理,并运用Risk Spectrum软件进行定量分析,得出系统失效概率约为9.215×10^-5。结果表明,余热排出热交换器入口管线上电动阀失效关闭是导致系统失效的最主要因素。基于计算结果提出两种方法对系统进行优化：1） 在另一回路增加同样1套PRHR系统;2） 在原有系统基础上增加一PRHR热交换器。通过故障树分析计算,并分别从系统可靠性、复杂性、经济性等方面对两种方法进行比较发现,方法2更具可行性,建议工程上采用此方法对系统进行优化。     

Study on construction of loss of component cooling water system initiating event fault tree for fire event PSA

ABSTRACT

In this study, the construction of the loss of component cooling water system (LOCCWS) initiating event (IE) fault tree (FT) for an actual fire event probabilistic safety assessment (PSA) model of the Korean reference nuclear power plant considering only IE initiators was validated. The quantification results of the LOCCWS accident sequences obtained using an LOCCWS IE FT model with only initiators are similar to that with initiators and enabling events. This confirmed that the LOCCWS IE FT for an actual fire event PSA model could be constructed by considering only IE initiators. In addition, the same LOCCWS accident sequences were quantified assuming that fire triggering only the LOCCWS IE leads to reactor shutdown. Compared with the quantification result obtained based on the assumption that any fire included in the fire event PSA leads to reactor shutdown, the core damage frequency (CDF) can be reduced. Thus, it can be concluded that there is a possibility of underestimation of CDF when the LOCCWS IE FT model with only initiators is used and the assumption that fire triggering only the LOCCWS IE results in reactor shutdown is employed for the quantification of LOCCWS accident sequences.     

备用系统可靠性动态问题方法探究

随着核电安全性日益受到世界公众关注,用概率论的方法对核电厂系统进行可靠性分析也越来越显示出其重要作用.在进行系统可靠性分析时,故障树方法是国际上公认的一种简单、有效、经典的方法.但随着所要分析的工程系统日趋庞大,系统工作过程日趋复杂,对含有动态过程的系统可靠性分析成为故障树方法中的棘手问题.本文以可靠性分析中的典型动态问题--备用系统的可靠性为研究对象,运用蒙特卡罗方法编程探究如何运用故障树来求解可靠性动态问题,并在进行程序验证后对一个备用系统实例进行计算,给出计算结果.     

Knowledge-based software design for Defense-in-Depth risk monitor system and application for AP1000

As part of the new risk monitor system, the software for the plant Defense-in-Depth (DiD) risk monitor system was designed based on the state-transition and finite-state machine, and then the knowledge-based software was developed by object-oriented method utilizing the Unified Modeling Language (UML). Currently, there are mainly two functions in the developed plant DiD risk monitor software that are knowledge-base editor which is used to model the system in a hierarchical manner and the interaction simulator that simulates the interactions between the different actors in the model. In this paper, a model for playing its behavior is called an Actor which is modeled at the top level. The passive safety AP1000 power plant was studied and the small-break loss-of-coolant accident (SBLOCA) design basis accident transient is modeled using the plant DiD risk monitor software. Furthermore, the simulation result is shown for the interactions between the actors which are defined in the plant DiD risk monitor system as PLANT actor, OPERATOR actor, and SUPERVISOR actor. This paper shows that it is feasible to model the nuclear power plant knowledge base using the software modeling technique. The software can make the large knowledge base for the nuclear power plant with small effort.     

Methodology for reliability allocation based on fault tree analysis and dualistic contrast

Reliability allocation is a difficult multi-objective optimization problem.This paper presents a methodology for reliability allocation that can be applied to determine the reliability characteristics of reactor systems or subsystems.The dualistic contrast,known as one of the most powerful tools for optimization problems,is applied to the reliability allocation model of a typical system in this article.And the fault tree analysis,deemed to be one of the effective methods of reliability analysis,is also adopted.Thus a failure rate allocation model based on the fault tree analysis and dualistic contrast is achieved.An application on the emergency diesel generator in the nuclear power plant is given to illustrate the proposed method.     

基于概率因果模型与遗传算法的船舶核动力装置故障诊断的新方法

介绍一种将概率因果模型和遗传算法相结合的核动力装置二回路凝给水系统的故障诊断方法,它将概率因果模型的似然函数作为遗传算法的适应函数,从而将复杂系统的故障诊断转化为最优问题。仿真结果表明,该方法能够适应诊断过程中出现的不确定性,并实现多故障诊断,具有较高的诊断可靠性和实用性。     

A dynamic neural network based accident diagnosis advisory system for nuclear power plants

In this work, an accident diagnosis advisory system (ADAS) using neural networks is developed. In order to help the plant operators quickly identify the problem, perform diagnosis and initiate recovery actions ensuring the safety of the plant, many operator support systems and accident diagnosis systems have been developed. The ADAS is a kind of such accident diagnosis system, which makes the task of accident diagnosis easier, reduces errors, and eases the workload of operators by quickly suggesting likely accidents based on the highest probability of their occurrence. In order to perform better than other accident diagnosis systems, the ADAS has three main objectives. To satisfy these three objectives, two kinds of neural networks that consider time factors are used in this work. A simple accident diagnosis system is implemented in order to validate the ADAS. After training the prototype, several accident diagnoses were performed. The results show that the prototype can detect the accidents correctly with good performance.     

基于朴素贝叶斯的核电厂事故诊断方法研究

本文将机器学习领域的贝叶斯技术应用于核应急中的电厂状态诊断,提出了基于朴素贝叶斯分类器的核电厂事故诊断方法。利用压水堆核电厂仿真机获取事故案例数据,对朴素贝叶斯分类模型进行训练,实现了对核电厂多类事故(LOCA、SGTR、MSLB)的诊断。测试结果表明,基于朴素贝叶斯分类器的核电厂事故诊断方法在诊断精度、诊断效率、事故类型可扩展性以及程序自主化诊断上有显著优势,并且模型训练中不同事故类型先验分布对诊断结果影响较小,具有较好的适用性。     

Fault diagnosis algorithm for 3-phase passive rectifiers based on frequency-domain analysis for acceleration grid power supply in CFETR NBI system

The acceleration grid power supply (AGPS) is a crucial part of the Negative-ion Neutral Beam Injection system in the China Fusion Engineering Test Reactor, which includes a 3-phase passive (diode) rectifier. To diagnose and localize faults in the rectifier, this paper proposes a frequency-domain analysis-based fault diagnosis algorithm for the rectifier in AGPS. First, time-domain expressions and spectral characteristics of the output voltage of the TPTL-NPC inverter-based power supply are analyzed. Then, frequency-domain analysis-based fault diagnosis and frequency-domain analysis-based sub-fault diagnosis algorithms are proposed to diagnose open circuit (OC) faults of diode(s), which benefit from the analysis of harmonics magnitude and phase-angle of the output voltage. Only a fundamental period is needed to diagnose and localize exact faults, and a strong Variable-duration Fault Detection Method is proposed to identify acceptable ripple from OC faults. Detailed simulations and experimental results demonstrate the effectiveness, quickness, and robustness of the proposed algorithms, and the diagnosis algorithms proposed in this article provide a significant method for the fault diagnosis of other rectifiers and converters.