Research on Wear life Prediction of CRDM Transmission Pair Based on Machine Learning
-
摘要: 控制棒驱动机构（CRDM）是反应堆内唯一具有相对运行的设备单元，可快速调节反应堆反应性，对反应堆安全运行十分重要，磨损是影响CRDM传动副功能失效的主要因素，直接决定其使用寿命。本文通过CRDM传动副磨损实验，发现传动副主要磨损形式有3种：磨粒磨损、疲劳磨损和氧化磨损，同时发现当传动副顶部区域磨损体积比达16.46%时，CRDM出现滑棒，可判定此刻转动部件出现了磨损失效，将此刻的磨损体积值作为传动副的失效阈值。在获得传动副磨损体积数据和外部振动信号后，本文构建了内部磨损体积与外部振动信号的关联关系，并通过外部振动信号，基于支持向量回归（SVR）、卷积神经网络（CNN）、长短期记忆网络（LSTM）3种机器学习算法，分别构建了CRDM传动副寿命预测模型，通过对比分析认为，在预测精度上LSTM模型优于CNN模型优于SVR模型，在计算效率上SVR模型优于CNN模型优于LSTM模型。
-
关键词:
- 控制棒驱动机构（CRDM） /
- 传动副 /
- 磨损 /
- 剩余寿命
Abstract: Control Rod Drive Mechanism (CRDM) is the only equipment unit with relative operation in the reactor. It can adjust the reactivity of the reactor quickly, so it is very important for the safe operation of the reactor. Wear is the main factor affecting the functional failure of CRDM transmission pair, directly determining its service life. Through the wear life test of the transmission pair of CRDM, it is found that the three main wear forms of the transmission pair are abrasive wear, fatigue wear and oxidation wear. At the same time, it is found that when the wear volume ratio at the top of the transmission pair reaches 16.46%, the sliding rod appears in the driving mechanism, which can be judged that the rotating parts have worn out, and the wear volume value at this moment is taken as the failure threshold of transmission pair. After obtaining the data of transmission pair wear degradation and external vibration signals, the relationship between internal wear and external vibration signals is constructed. Through external vibration signals and based on three machine learning algorithms SVR, CNN and LSTM, the life prediction models of CRDM transmission pair were established respectively. Through comparative analysis, it is concluded that in terms of prediction accuracy, the LSTM model outperforms the CNN model, which in turn outperforms the SVR model, while in terms of computational efficiency, the SVR model surpasses the CNN model, which in turn surpasses the LSTM model.-
Key words:
- Control rod drive mechanism (CRDM) /
- Transmission pair /
- Wear /
- Residual life
-
图 3 传动副在不同应力水平下磨损量退化曲线
Figure 3. Wear Degradation Curves of Transmission Pair under Different Stress Levels
图 5 传动副转动部件磨损区域的微观形貌与不同点的元素占比
Figure 5. Microscopic Morphology of Wear Areas of Rotating Components in Transmission Pairs and Proportion of Elements at Different Points
图 6 传动副振动信号频域曲线
Figure 6. Frequency Domain Curve of Vibration Signal of Transmission Pair
图 7 信号特征的皮尔逊相关性系数曲线
Figure 7. Pirsson Correlation Coefficient Curve of Signal Characteristics
表 1 加速应力水平划分
Table 1. Classification of Accelerated Stress Levels
加速应力水平 轴向载荷（F）/N 棒速（V）/(mm·s−1) Y0 700 10 Y1 1000 10 Y2 1500 10 Y3 2000 10 
下载: 导出CSV
表 2 SVR模型评价指标
Table 2. SVR Model Evaluation Indicators
评价指标 训练集 测试集 RMSE 29.7895 46.1547 MAE 11.6587 17.1589 CRA/% 68.689 59.137 RMSE—均方根误差；MAE—平均绝对误差；CRA—累积相对计算精度
下载: 导出CSV
表 3 CNN模型评价指标
Table 3. CNN Model Evaluation Indicators
评价指标 训练集 测试集 RMSE 17.5663 27.4624 MAE 5.2466 8.2697 CRA/% 87.342 76.594
下载: 导出CSV
表 4 LSTM模型评价指标（滑动窗口数为10）
Table 4. Evaluation Indicators of the LSTM Model with a Sliding Window of 10
评价指标 训练集 测试集 RMSE 11.0564 19.6658 MAE 3.1587 5.9786 CRA/% 92.109 83.768
下载: 导出CSV
表 5 3种模型计算精度和效率对比表
Table 5. Comparison of Accuracy and Efficiency among the Three Models
预测模型 CRA/% 泛化处理时间/s SVR 59.137 1.53 CNN 76.594 10.17 LSTM 83.768 15.24
下载: 导出CSV
-
[1] 杨云,于天达,陈杨明,等. 控制棒驱动机构磁路饱和特性分析[J]. 核动力工程,2023, 44(3): 138-143. [2] 谢细明,彭航,张倬,等. 核反应堆控制棒驱动机构动作部件状态检测研究[J]. 核动力工程,2019, 40(2): 90-94. [3] 于天达,张进强,罗亮,等. 控制棒驱动机构钩爪齿复合冲击磨损可靠性分析[J]. 机械设计,2021, 38(4): 16-22. [4] 于天达,张志强,刘彦霆,等. 控制棒驱动机构销轴磨损寿命与可靠性仿真[J]. 兵器装备工程学报,2021, 42(3): 219-224. doi: 10.11809/bqzbgcxb2021.03.040 [5] 刘世航,李经纬,张冠华,等. 池式供热堆用控制棒驱动机构钩爪磨损寿命分析[C]// 中国核学会. 中国核科学技术进展报告(第八卷)中国核学会2023年学术年会论文集 第9册 核安全 核设备 反应堆热工流体力学. 北京: 中国原子能科学研究院,2023. [6] 周军,陈勇,罗强,等. CRDM钩爪用stellite-6合金冲击磨损性能研究[J]. 核动力工程,2016, 37(3): 66-69. [7] 李玲,王丰,王永东. 控制棒驱动机构钩爪摩擦副实验[J]. 理化检验(物理分册),2016, 52(6): 384-388. [8] 李维,杨方亮,杨晓晨,等. 控制棒驱动机构传动件耐磨性研究[J]. 科技视界,2016(12): 25-26. doi: 10.3969/j.issn.2095-2457.2016.12.013 [9] 郭学卫,申永军,杨绍普. 基于样本熵和分数阶傅里叶变换的滚动轴承故障特征提取[J]. 振动与冲击,2017, 36(18): 65-69. [10] 董立朋,聂清浩,孙晓坤,等. 基于皮尔逊相关系数法的盾构掘进参数对地表沉降影响分析[J]. 施工技术(中英文),2024, 53(1): 116-123. [11] AKHAND R, PRIYAV K, S P B, et al. Gear fault diagnosis based on complex network theory and error-correcting output codes: Multi class support vector machine[J]. Proceedings of the Institution of Mechanical Engineers, 2024, 238(6): 1135-1152. [12] ATUL D, BHAGAT S, PANKAJ G. Diagnosing Faults in Rolling Bearings of an Air Compressor Set Up Using Local Mean Decomposition and Support Vector Machine Algorithm[J]. Journal of Vibration Engineering & Technologies, 2024, 12(4): 6635-6648. [13] MA Y, WANG Z, GAO J, et al. A novel method for remaining useful life of solid-state lithium-ion battery based on improved CNN and health indicators derivation[J]. Mechanical Systems and Signal Processing, 2024, 220:111646-111646. [14] KUMAR P S, NANDAN R R. LSTM-based deep learning approach for remaining useful life prediction of rolling bearing using proposed C-MMPE feature[J]. Journal of Mechanical Science and Technology, 2024, 38(5): 2197-2209. -
计量
- 文章访问数: 30
- HTML全文浏览量: 6
- PDF下载量: 3
- 被引次数: 0
