Abstract

This paper proposes the application of a pattern recognition–based technique to enhance the process of control rod position identification in pressurized water reactors (PWRs). The proposed technique employs a multivariant analysis technique, namely, principal component analysis (PCA) and clustering analysis (CA) to identify the position of the PWR control rod using its impact on the core radial thermal neutron flux along the axial track of motion. The results of these investigations have shown that the proposed technique successfully removed the limitation on the data size and any limitations imposed by outlier samples, extracted the noise, and provided near-instantaneous analytical and visual ways for position identification process with excellent generalization fitting and prediction efficiencies. In the context of this paper, multiple in-depth simulations are conducted to ascertain the efficiency of the proposed technique in identifying the control rod positions. These simulations have been conducted using a Westinghouse 2772-MW(thermal) PWR benchmark at 100% thermal power generation, where a three-dimensional TRITON FORTRAN-code has been utilized to simulate the radial thermal neutron flux of the PWR core. The PCA model is developed, tested, and generalized using the SIMCA software package. In addition, CA is also performed via the Minitab statistics software package in order to confirm the efficiency of the proposed technique.

摘要

本文提出了一种基于模式识别的技术的应用，以增强压水堆（PWR）中控制杆位置的识别过程。所提出的技术采用多变量分析技术，即主成分分析（PCA）和聚类分析（CA），以利用PWR控制棒对沿轴向运动轨迹的芯径向热中子通量的影响来确定PWR控制棒的位置。这些研究的结果表明，所提出的技术成功地消除了数据大小的限制和异常样本所施加的任何限制，提取了噪声，并为位置识别过程提供了近乎瞬时的分析和视觉方式，具有出色的泛化拟合和预测能力。效率。在本文的背景下，进行了多次深入模拟，以确定所提出技术在识别控制杆位置方面的效率。这些模拟是在100％火力发电时使用Westinghouse 2772兆瓦（热）PWR基准进行的，其中三维TRITON FORTRAN代码已用于模拟PWR堆芯的径向热中子通量。PCA模型是使用SIMCA软件包开发，测试和推广的。此外，还通过Minitab统计软件包执行CA，以确认所提出技术的效率。其中使用了三维TRITON FORTRAN代码来模拟PWR堆芯的径向热中子通量。PCA模型是使用SIMCA软件包开发，测试和推广的。此外，还通过Minitab统计软件包执行CA，以确认所提出技术的效率。其中使用了三维TRITON FORTRAN代码来模拟PWR堆芯的径向热中子通量。PCA模型是使用SIMCA软件包开发，测试和推广的。此外，还通过Minitab统计软件包执行CA，以确认所提出技术的效率。