基于模糊距离的核电厂瞬态分段识别方法

近年来发展的核电厂瞬态识别技术,可为操纵员提供处于发展阶段的故障信息,有助于了解核电厂状态并及时采取相应的操作动作,保证核电厂的安全运行。将瞬态过程曲线分为两段,前段利用聚类方法用于快速识别,后段利用提取的瞬态过程的特征进行更准确的识别。利用待识别瞬态与参考瞬态间的模糊距离描述二者的相近程度,可以消除噪声等扰动的影响,并得到更符合认知习惯的结果。利用高温气冷堆核电厂仿真机的故障数据验证瞬态识别方法的有效性。     

基于SVM的核电厂瞬态识别可行性研究

SVM（Suppon Vector Machine，向量支持机）是根据统计学理论研究得到的一种识别事故的新机器。它对核电厂的各种事故瞬态行为的识别具有巨大潜力，得到了广泛的研究。用于识别核电厂事故的计算机支持系统的主要任务就是瞬态识别，识别核电厂运行状态的能力至关重要。为此，工作人员可以选择适当的响应动作来完成。基于模数的第四版事故分析程序可以用于模拟核电厂中各种正常和异常事件。本文通过MAAP程序来描述沸水堆中各种冷却剂丧失事故。这些事故中，传感器的输出被用于测试SVM事故识别机。SVM的计算结果表明，它们可以作为一种较好的分类机应用于瞬态识别。     

瞬态监督和处理方法在核电厂运行许可证延续中的应用

核电厂延续运行前,由于缺乏瞬态监督管理对核电厂延续寿命影响的具体认知,相关瞬态监督和控制仅限于设计寿期范围内,没有涉及到延续运行。这导致核电厂在运行前期没有针对性地管理瓶颈瞬态的消耗,从而减少了核电厂实际可达的寿命长度;或者相关瞬态数据收集不够详细,不足以支撑更细致的疲劳分析,在延续运行评估时只能采取更多的包络处理,难以实现更长的评估寿命。本文针对上述此问题,通过汲取秦山核电厂延续运行研究中瞬态相关经验,从日常运行监督和专项延续评估两个方面,对核电厂延续运行瞬态监督和数据处理进行研究,形成了适用于核电厂延续运行的瞬态管理技术方法,可有效指导后续核电机组开展延续运行工作。     

采用人工神经网络的基于征兆的核电厂运行诊断系统

核电厂运行过程中有许多瞬态过程，装置失灵、处理系统故障和安全系统失效都可能引起瞬态过程。这些瞬态过程可能导致电厂进入非期望的异常状态。如果电厂处于非期望状态(一般称为始发事件)，操纵员必须执行诊断和纠正措施，但是操纵员的反应可能太慢而不能减轻故障的不利影响。而本论文的目的是开发一种基于人工神经网络的操纵员支持系统，这个系统将辅助操纵员在始发事件的发展初期将其识别出来。电厂异常情况可通过处理仪表的读数来诊断和识别。基于征兆的诊断系统用于识别始发事件。该系统对始发事件的识别是通过使用弹性反传算法(Resilient Back Propagation Neural Network Algorithm)来实现。一旦识别出始发事件，系统将显示始发事件类型和必要的操纵员规程，还显示相关参数的趋势曲线。目前该系统能够识别Narora核电站的8个始发事件。本文将以一个始发事件为例，说明该诊断系统的特征。     

核电厂汽轮机详细数值建模研究及其瞬态分析

以岭澳一期核电厂汽轮机部件为原型,利用系统程序RELAP5对其进行详细数值建模研究。通过在100%功率稳态工况下的计算证明,详细的汽轮机数值建模弥补了简化建模中焓值计算误差较大的缺陷。将详细的汽轮机数值建模整合到全范围核电厂热力系统模型中进行瞬态分析,并与岭澳一期核电厂原始实验报告中汽轮机负荷从97%功率水平阶跃变化至87%功率水平瞬态运行工况的数据曲线进行对比。结果表明,稳态模型的焓计算值与电厂实际值误差在2%以内,瞬态模型的分析参数趋势符合电厂实际情况。     

百万千瓦级核电厂电网故障甩负荷到厂用电瞬态研究

在核电厂机组运行时，如果母线断路器找开，机组与外电网失去连接，交流发电机继续向机组厂用设备供电，该瞬态称为甩负荷到厂用电，由于电网故障的原因。机组保护系统自动找开母线断路器，称为电网故障甩负荷。甩负荷到厂用电运行可以提高机组的可用性和运行的安全性。本文采用CATIA2－A程序分析了百万千瓦级核电厂在电网故障导致甩负荷到厂用电运行的瞬态变化，由于寿期初、末的反应性系数不同，会导致中子通量峰值的不同，从而决定了瞬态能否成功。     

基于模糊熵的核电站瞬态识别方法

为保障核电站安全经济运行,需及时准确地识别核电站出现的异常。本文通过处理关键变量的时间序列数据,对瞬态过程进行识别：利用模糊熵度量时间序列复杂度的能力,判断系统是否处于正常状态;进而利用互模糊熵度量两时间序列相似度的能力,对出现的瞬态进行类型识别。利用模块式高温气冷堆核电站仿真机的数据验证了本方法的可行性和有效性,结果表明本文方法可有效进行瞬态识别,且不需复杂的训练过程。     

秦山300MW核电机组全范围仿真机故障和瞬态测试

简要介绍了秦山３００ＭＷ核电机组全范围仿真机验收测试规程中故障和瞬态部分的主要内容，并提出了在仿真机上实现在故障和瞬态的方法和测试手段，同时介绍了仿真机故障测试的主要过程和结果。     

瞬态曲线分段线性化方法研究及其程序实现

核电厂的瞬态数据是相关设备应力分析和评定的必要输入。然而,电厂实际运行记录数据或设计瞬态计算数据通常由时间间隔很短的大量数据点构成,直接作为应力分析的输入将会导致大量不必要的计算。为了简化分析,一般需要将瞬态数据曲线进行分段线性化处理。本文基于最小二乘法,开发了一种瞬态曲线分段线性化的方法,并且通过编写PLTC(PiecewiseLinearization forTransientCurves)程序予以实现。分别采用标准正弦函数曲线和核电厂典型瞬态曲线对本文的方法进行了验证,结果表明,本文的方法和程序能够很好地实现瞬态曲线的分段线性化。本文的研究有助于提高核电厂瞬态曲线分段线性化的效率和精度。     

基于KNN方法的熔盐堆系统瞬态识别模型开发及分析

熔盐堆作为第四代先进核能系统之一，在安全性、经济性、防核扩散和可持续性等方面具有独特的优势。为了保障熔盐堆运行安全，需要快速、准确地识别瞬态工况，目前的瞬态识别方法主要依赖于操作员人工识别，这会引入较大的人为因素，严重影响核电安全。为了减少熔盐堆系统瞬态识别过程中引入的人为因素，提高熔盐堆运行安全，使用RELAP5-TMSR程序对美国橡树岭国家实验室建造运行的熔盐实验堆（Molten Salt Reactor Experiment,MSRE）的瞬态工况进行建模与仿真，产生数据集，基于K近邻（K-Nearest Neighbor,KNN）机器学习方法，建立了熔盐堆系统瞬态识别模型，并对识别模型在噪声下的鲁棒性进行了分析和优化。结果显示：基于KNN方法建立的熔盐堆系统瞬态识别模型在测试集上的F1分数达到99.99%；在噪声下的识别F1分数达到94.32%，具有较高的鲁棒性；进一步优化后的熔盐堆系统瞬态识别模型在噪声下的F1分数达到99.73%，能较为准确地识别MSRE的瞬态工况，满足熔盐堆系统瞬态识别需求。基于KNN方法的熔盐堆系统瞬态识别模型能够有效识别系统瞬态工况，可应用于熔盐堆智能运...     

Transient Boiling of Sodium in Seven-Pin Bundle under Transient Overpower Conditions

Transient sodium boiling experiments were conducted in an electrically heated 7-pin bundle under transient overpower conditions. In each run the heater power was gradually raised at almost constant rate under forced convection.

The observed coolant voiding was initially limited to the central subchannel on account of an appreciable time lag in temperature rise occurring between the central and peripheral subchannels. This would appear to call for calculations with two-dimensional voiding model.

The bulk pressure rises registered upon initial vaporization were markedly lower than the vapor pressure corresponding to the incipient-boiling (IB) wall superheat. The pressure pulse generated upon vapor bubble collapse correlated reasonably well with the re-entrant liquid velocity, but the measured value was very much smaller than predicted theoretically from sodium hammer analysis.     

核主泵变流量过渡过程瞬态水力特性研究

为研究核主泵从设计工况向非设计工况过渡过程的瞬态水力特性及内部流动机理,应用计算流体力学软件CFX对核主泵叶轮流道内的变流量瞬态流动特性进行数值模拟计算。研究结果表明：变流量过渡时,核主泵的压力脉动沿圆周方向分布并不均匀,其变化趋势是逐渐上升到最大值后又降低,基本呈正弦变化规律,瞬态压力波动变化次数等于叶片与导叶片数之间的动静干涉次数,监测点越靠近叶片与导叶交界面,压力波动越大;由于冲角的存在造成叶轮流道内的速度呈先下降后上升的变化趋势;导叶不仅具有将动能转换为压能的功能,同时也具有有效减缓压力脉动幅度的功能;向小流量过渡时,由于流量减少,在靠近叶轮出口处出现二次回流,造成叶轮流道内速度变化幅度随流量的减少而增大。     

Transient of a Pulse Reactor

Power and temperature transients following a prompt critical pulse applied to an experimental reactor has been studied by prompt jump approximation in two-dimensional phase space. The phenomena are classified into several types characterized by the amount of inserted reactivity and the value of cooling time constant. The temperature is shown to exceed in no case a level twice as high as the equilibrium value.     

Transient failure of zircaloy cladding

A state of the art study of transient failure of zircaloy reveals that a wide range of failure models exist, and their use in fuel analysis depends upon the type of application. Licensing analyses, for example, employ highly simplified criteria that depict cladding rupture in terms of a single response variable, usually the nominal circumferential stress, as function of temperature; while fuel behavior analyses use more detailed models with varying levels of sophistication. Some of these models, however, are ill-suited for detailed fuel rod analysis computer codes because of their dependence on case-specific parameters; others are limited in their utility because of inherent deficiencies in their theoretical treatment of the phenomena involved. Based on this study, a new zircaloy transient failure model is developed that takes account of the important variables involved in transient phenomena, namely, heating rate, strain rate, temperature, true stress, true strain, and prior condition of the cladding material. The model is particularly suited for detailed fuel rod analysis and thus was implemented in the FREY code.     

Mechanisms of Transient Radiation Effects

Transient radiation effects are defined to be manifestations of electrons excited in materials, i. e., ionization effects. Manifestations of particular interest are the emission of secondary electrons from surfaces, conduction in ionized gases, semiconductors and insulators, and the emission and absorption of optical radiation by excess electrons. In each case the processes can be described as a sequence leading from the primary interaction of the incident nuclear radiation with an atom in the target material through the slowing down of the electrons which are detached in this interaction to the eventual recapture or trapping of the free electrons. Most of the phenomena can be described in terms of a mean lifetime for the electrons and a mobility. Assuming reasonable values for these quantities, calculations can be performed to estimate the order of magnitude of transient effects in typical materials and electronic components.     

Transient nonhomogeneous equilibrium water-steam-air flow

A mathematical model of transient equilibrium inhomogeneous two-phase two-component water-steam-air flow has been developed. The idea of a quasicontant slip is used. The set of four differential equations is transformed into non-conservative form, so that it can be used directly to build the numerical solution in the time space domain with highly effective numerical methods. From the steady-state system an expression is found defining the critical mass flow rate. There is a good agreement with the existing theories for the limiting cases of a missing phase of component.     

Transient analysis of DTT rakes

This paper presents an analytical model for the determination of the cross-sectionally averaged transient mass flux of a two-phase fluid flowing in a conduit instrumented by a Drag-Disk Turbine Transducer (DTT) Rake and a multibeam gamma densitometer.Parametric studies indicate that for a typical blowdown transient, dynamic effects such as rotor inertia can be important for the turbine-meter. In contrast, for the drag-disk, a frequency response analysis showed that the quasisteady solution is valid below a forcing frequency of about 10 Hz, which is faster than the time scale normally encountered during blowdowns.The model showed reasonably good agreement with full scale transient rake data, where the flow regimes were mostly homogeneous or stratified, thus indicating that the model is suitable for the analysis of a DTT rake.     

Transient Nonlinear Response of Plastic Scintillators

The peak dose rate for which light output is linearly proportional to incident radiation intensity was measured by bombarding 80-mil-thick samples of NE 102 and MEL 150C scintillators with 12-MeV electrons from the EG&G/AEC linear accelerator. To obtain the equivalent gamma dose rate, the sensitivity of the scintillator-photodiode detector used in this work was measured with a calibrated Co60 source. Nonlinear behavior was observed for excitation pulse widths between six nanoseconds and 0.5 microseconds. Analysis of the detector signals indicates that nonlinearity is a function of total dose as well as dose rate. Thus for an excitation pulse width of 0.5 microseconds, scintillator output was found to be linear at the start of the burst, but showed marked nonlinearity by the end of the pulse. The data also indicate that if the dose rate drops below the threshold value after a scintillator is driven into nonlinear operation, the scintillator will continue to behave nonlinearily for some time. The decay time associated with recovery from nonlinearity is greater than 0.5 microsecond. Data are presented, showing the average dose rate required to produce 5% and 10% nonlinear behavior as a function of exposure time. As a typical case, 10% nonlinear behavior for NE 102 results from an average dose rate of 6.5 × 1011 R/sec for an exposure time of 10 nanoseconds.     

Transient thermal-hydraulic analysis for reactor cores

A new computational method is presented for a transient, thermal-hydraulic, multichannel analysis. The method is developed based on the concept of artificial compressibility to preserve the elliptic character of the reactor core flow in order to satisfy the realistic pressure boundary conditions, and to account for the discontinuities of the emprical correlations simulating the flow resistances. The computer code (RETSAC) developed by implementing the method presented in this paper can be categorized as a fourth generation multichannel computer code. This new computer code has been compared with the widely used marching techniques, such as COBRA IIIC (the third generation). The numerical results clearly indicate the situations in which the marching technique may or may not be appropriate. Furthermore, the RETSAC computer code can calculate various normal or off-normal reactor core flows which the third generation codes could not handle without a substantial increase of computer time.