堆芯燃料棒自动校对算法研究

李向东; 蒋和松; 王学渊; 胥学金; 何小春

doi:10.13832/j.jnpe.2023.04.0203

堆芯燃料棒自动校对算法研究

doi: 10.13832/j.jnpe.2023.04.0203

李向东^1,,
蒋和松^{1, 2},
王学渊^{1, 2, ,},
胥学金¹,
何小春³

1.
西南科技大学信息工程学院，四川绵阳，621000
2.
特殊环境机器人技术四川省重点实验室，四川绵阳，621000
3.
四川九洲空管科技有限公司，四川绵阳，621000

基金项目: 国家自然科学基金（61771411）；四川省教育厅项目（JG2021-871）

详细信息

作者简介:
李向东（1996—），男，硕士研究生，现主要从事反应堆燃料组件校对及图像拼接方面研究，E-mail: 953716816@qq.com

通讯作者:
王学渊，E-mail: 121053406@qq.com

中图分类号: TL36;TL38
计量
- 文章访问数: 113
- HTML全文浏览量: 37
- PDF下载量: 21
- 被引次数: 0
出版历程
- 收稿日期: 2022-09-02
- 修回日期: 2023-04-29
- 刊出日期: 2023-08-15

Research on Automatic Calibration Algorithm of Reactor Fuel Rods

Li Xiangdong^1
,,
Jiang Hesong^{1, 2},
Wang Xueyuan^{1, 2
, ,},
Xu Xuejin¹,
He Xiaochun³

1.
School of Information Engineering, Southwest University of Science and Technology, Mianyang, Sichuan, 621000, China
2.
Robot Technology Used for Special Environment Key Laboratory of Sichuan Province, Mianyang, Sichuan, 621000, China
3.
Sichuan Jiuzhou Aerocont Technologies Co., Ltd., Mianyang, Sichuan, 621000, China

摘要

摘要: 由于核反应堆经常更换燃料棒，为保证反应堆的安全运行，需准确确定堆芯燃料棒的类别及其安装的位置。该校对算法先根据堆芯燃料棒安装位置的分布关系建立全局和局部虚拟二维坐标映射模型。识别时，拍摄各视点局部序列图，识别局部图中燃料棒的中心位置，并标定局部虚拟二维坐标映射模型，再利用燃料棒的中心位置与标定后的局部映射模型中的位置做欧氏距离判断，实现类别重构，进一步得到堆芯全景拼接图，以辅助校对。仿真研究表明，本算法可以有效检测燃料棒的类别及安装位置，识别率高于98%，准确率达到100%，全景拼接结果稳定可靠。在堆芯燃料棒的校对工作中具有较大的应用潜力。
- 燃料棒 /
- 多视角 /
- 特征匹配 /
- 配准 /
- 深度学习
Abstract: Since the nuclear reactors need frequent replacement of fuel rods, it is necessary to determine the type and installation position of the core fuel rods accurately to ensure the safe operation of the reactor. Herein, the global and local virtual two-dimensional coordinate mapping models have been established in terms of the distribution relationship of fuel rod installation positions. The local sequence pictures of each viewpoint are taken to identify the central position of the fuel rods in the local pictures, and the local virtual two-dimensional coordinate mapping model is calibrated. Then, the Euclidean distance between the central position of the fuel rods and the position in the calibrated local mapping model is measured to realize type refactor, and the core panoramic mosaic is further obtained to assist calibration. The simulation results show that the algorithm can effectively detect the type and installation position of fuel rods, the recognition rate is higher than 98%, the accuracy rate reaches 100%, and the panoramic mosaic results are stable and reliable. It has great application potential in the calibration of core fuel rods.
- Fuel rods /
- Multi-view /
- Feature matching /
- Registration /
- Deep learning

HTML全文

图 1 多帧识别结果聚类

DBSCAN——基于密度的空间聚类算法

Figure 1. Clustering of Multi Frame Recognition Results

下载: 全尺寸图片幻灯片

图 2 多帧识别结果聚类及清晰帧选取流程

Figure 2. Multi Frame Recognition Result Clustering and Clear Frame Selection Process

下载: 全尺寸图片幻灯片

图 3 左视角燃料棒识别结果图

Figure 3. Recognition Result of Fuel Rods from Left View Angle　　　　　　　

下载: 全尺寸图片幻灯片

图 4 全局虚拟二维坐标映射模型及相机覆盖区域

Figure 4. Global Virtual 2D Coordinate Mapping Model and Camera Coverage Area

下载: 全尺寸图片幻灯片

图 5 左视角局部虚拟二维坐标映射模型标定结果图

Figure 5. Calibration Results of Local Virtual 2D Coordinate Mapping Model from Left View Angle

下载: 全尺寸图片幻灯片

图 6 燃料棒类别重构原理图

Figure 6. Schematic Diagram of Fuel Rod Type Refactor

下载: 全尺寸图片幻灯片

图 7 堆芯全景拼接图

Figure 7. Panoramic Mosaic of Core

下载: 全尺寸图片幻灯片

图 8 不同方法特征点匹配对数

Figure 8. Matching Logarithms of Feature Points for Different Methods

下载: 全尺寸图片幻灯片

图 9 不同方法特征点正确匹配对数

Figure 9. Correct Matching Number of Feature Point Pairs of Different Methods

下载: 全尺寸图片幻灯片

图 10 不同方法耗时

Figure 10. Time Consumption of Different Methods

下载: 全尺寸图片幻灯片

图 11 不同方法配准精度

Figure 11. Registration Accuracy of Different Methods

下载: 全尺寸图片幻灯片

表 1 10组堆芯模拟体及堆芯现场局部图类别重构结果

Table 1. Type Refactor Results for 10 Sets of Core Simulators and Core Site Local Pictures

实验组号		燃料棒总数	单帧YOLO v5 类别重构		本文方法
实验组号		燃料棒总数	识别率/%	准确率/%	识别率/%	准确率/%
堆芯模拟体	1	257	99.4	99.2	99.4	100
	2	245	99.6	99.4	99.4
	3	257	99.6	99.2	99.8
	4		99.2	98.2	99.8
	5		98.2	99.2	98.4
堆芯现场	1	259	97.4	98.2	99.6	100
	2		96.6	97.4	99.8
	3		95.6	96.2	99.8
	4		97.2	98.2	99.4
	5		96.7	97.4	99.8

下载: 导出CSV

表 2 10组局部图全景拼接的RMSE

Table 2. RMSE for Panoramic Stitching of 10 Sets of Local Pictures

实验组号		RMSE/像素
实验组号		中、下局部拼接	中、上局部拼接	中、左局部拼接	中、右局部拼接
堆芯模拟体	1	1.8	1.8	1.6	2.4
	2	2.4	1.9	1.7	2.5
	3	1.8	1.7	1.7	1.6
	4	2.1	1.8	1.9	1.6
	5	1.7	1.8	1.6	2.4
堆芯现场	1	2.1	2.3	1.8	2.4
	2	2.0	1.9	1.8	1.8
	3	1.6	1.6	1.6	1.7
	4	1.9	1.9	2.3	1.6
	5	1.6	1.9	1.8	1.9

下载: 导出CSV

参考文献(16)

[1]	余红星,李文杰,柴晓明,等. 数字反应堆发展与挑战[J]. 核动力工程,2020, 41(4): 1-7. doi: 10.13832/j.jnpe.2020.04.0001
[2]	李开宇,蔡琦,才鑫馨,等. 数字孪生技术在浮动核电站设计阶段中的应用研究[J]. 核动力工程,2022, 43(1): 197-201.
[3]	高祖瑛,张作义,董玉杰. 核动力系统模拟技术的研究[J]. 核动力工程,1998, 19(2): 178-183.
[4]	HE K M, ZHANG X Y, REN S Q, et al. Deep residual learning for image recognition[C]//Proceedings of 2016 IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas: IEEE, 2016: 770-778.
[5]	KE X, ZHANG X L, ZHANG T W, et al. SAR ship detection based on an improved faster R-CNN using deformable convolution[C]//Proceedings of 2021 IEEE International Geoscience and Remote Sensing Symposium IGARSS. Brussels, Belgium: IEEE, 2021: 3565-3568.
[6]	LOWE D G. Distinctive image features from Scale-Invariant keypoints[J]. International Journal of Computer Vision, 2004, 60(2): 91-110. doi: 10.1023/B:VISI.0000029664.99615.94
[7]	BAY H, TUYTELAARS T, VAN GOOL L. SURF: speeded up robust features[C]//Proceedings of the 9th European Conference on Computer Vision. Graz, Austria: Springer, 2006: 404-417.
[8]	RUBLEE E, RABAUD V, KONOLIGE K, et al. ORB: an efficient alternative to SIFT or SURF[C]//Proceedings of 2011 International Conference on Computer Vision. Barcelona, Spain: IEEE, 2011: 2564-2571.
[9]	吕禾丰,陆华才. 基于YOLOv5算法的交通标志识别技术研究[J]. 电子测量与仪器学报,2021, 35(10): 137-144.
[10]	ESTER M, KRIEGEL H P, SANDER J, et al. A density-based algorithm for discovering clusters in large spatial databases with noise[C]//Proceedings of the Second International Conference on Knowledge Discovery and Data Mining. Portland: AAAI Press, 1996: 226-231.
[11]	MAHMUD M N, OSMAN M K, ISMAIL A P, et al. Road image segmentation using unmanned aerial vehicle images and DeepLab V3+ semantic segmentation model[C]//Proceedings of the 11th IEEE International Conference on Control System, Computing and Engineering. Penang, Malaysia: IEEE, 2021: 315-320.
[12]	JI Q B, WANG L J, HOU C B, et al. SAR and optical image matching eased on phase congruency and template matching[C]//Proceedings of the 8th International Conference on Dependable Systems and Their Applications. Yinchuan, China: IEEE, 2021: 315-320.
[13]	LIAO F R, CHEN Y, CHEN Y P, et al. SAR image registration based on optimized RANSAC algorithm with mixed feature extraction[C]//Proceedings of 2020 IEEE International Geoscience and Remote Sensing Symposium. Waikoloa, USA: IEEE, 2020: 1153-1156.
[14]	ZARAGOZA J, CHIN T J, BROWN M S, et al. As-projective-as-possible image stitching with moving DLT[C]//Proceedings of 2013 IEEE Conference on Computer Vision and Pattern Recognition. Portland: IEEE, 2013: 2339-2346.
[15]	POURFARD M, HOSSEINIAN T, SAEIDI R, et al. KAZE-SAR: SAR image registration using KAZE detector and modified SURF descriptor for tackling speckle noise[J]. IEEE Transactions on Geoscience and Remote Sensing, 2022, 60: 5207612.
[16]	LI Z F, CHEN C. An improved adaptive threshold BRISK feature matching algorithm based on SURF[C]//Proceedings of 2018 Chinese Automation Congress. Xi’an, China: IEEE, 2018: 2928-2932.