Study on Early Warning of Small Leakage in Primary Loop Based on Data Mining
-
摘要: 一回路小泄漏的过度演变有可能引起严重事故，为防止事故发生，提出一种基于多特征参数综合的改进高斯混合模型-灰色关联度法-熵权法（GMM-GRA-EWM）的故障预警方法。首先，对一回路小泄漏的动态运行特性进行机理分析，确定了预警特征参数。然后，根据已确定的预警特征参数，结合熵权法和灰色关联度法，建立多参数综合预警模型。最后，采用相关性分析、改进高斯混合模型算法有效学习了大量数据的统计特性，使预警阈值在不同工况下具有自适应能力。结果表明，该方法在变工况运行条件下，可以有效达到预警。相较于单参数和固定阈值预警，该方法具有更好的稳定性，预警更加准确、有效、及时，可为实现一回路系统的状态监测提供参考。Abstract: An excessive small leakage in the primary loop may cause serious accidents. In order to prevent such accidents, this study proposes an fault early-warning method of the improved Gaussian mixture model (GMM)-grey relational analysis (GRA)-entropy weighting method (EWM) based on the integration of multiple characteristic parameters. In this method, firstly, the mechanism of the dynamic operating characteristics of the small leakage in the primary loop is analyzed, and the early-warning characteristic parameters are thus determined. Then, based on these characteristic parameters, in combination with the EWM and GRA, a multi-parameter integrated early-warning model is established. Finally, the relational analysis and improved GMM algorithm are used to enable effective learning of the statistical features of massive data so that the early-warning thresholds can be self-adapted to different conditions. As this study shows, effective early warning can be realized in this method under variable operating conditions. Compared with single parameter and fixed threshold warning, this method is more stable and provides more accurate, effective and timely warning, and thus, it can provide a reference for the condition monitoring of the primary system.
-
图 1 稳压器压力控制系统
ΔP—压力偏差；G(S)—电加热器功率变化时稳压器压力的动态特性；Gre—电加热器功率；Pz—稳压器压力测量值
Figure 1. Pressurizer Pressure Control System
图 2 稳压器液位控制系统
Hz—稳压器液位测量值；ΔH—液位偏差；G1(S)—上充流量变化时稳压器液位的动态特性；fu—上充流量测量值；furef—上充流量设定值；Δf—fu与furef产生的偏差；Rcv—上充阀门开度
Figure 2. Pressurizer Level Control System
图 3 冷却剂平均温度控制系统
Tav—冷却剂平均温度测量值；Pm—汽轮机负荷和最终功率设定值中的最大值；ΔPm—与核功率的偏差；F(X)—函数发生器；G2(S)—在控制棒棒速扰动下堆芯冷却剂平均温度的动态特性；Crq—核功率；Rs—控制棒棒速
Figure 3. Average Coolant Temperature Control System
图 4 改进的高斯混合模型算法
D(xt, m)—马氏距离，即新样例xt到当前m个高斯成分的差异；St—新样例邻域集合，${S_t}{\rm{ = }}\left\{ {m = 1,{\rm{ }}2,{\rm{ }}\cdots,{\rm{ }}M|D({x_t},m) < {T_i}} \right\}$；Ti是第i个高斯成分的相似度阈值；U—当前模型平均有效训练样本数量，$U = \displaystyle \sum\limits_{m = 1}^M {\dfrac{{{u_{i}}}}{m}}$；ui—第i个高斯成分样本的数量；β—常数
Figure 4. Algorithm for Improved GMM
表 1 预警特征参数
Table 1. Early-Warning Characteristic Parameters
序号 特征参数 符号 1 稳压器压力 Pz 2 稳压器液位 Hz 3 热段温度 Tout 4 冷段温度 Tin 5 一回路冷却剂流量 Qz 6 上充阀门开度 Rcv 7 电加热器功率 Gre 8 温度控制棒棒位 Rd 
下载: 导出CSV
表 2 Pearson相关系数
Table 2. Pearson Correlation Coefficient
Y Crq Crq Crq Crq Crq Crq Crq Crq X Pz Hz Tout Tin Qz Rcv Gre Rd Pearson相关系数 0.8439 0.9902 0.9968 0.85 −0.9805 −0.3802 0.5923 0.72
下载: 导出CSV
表 3 不同工况下预警特征参数的均值和方差
Table 3. Mean Values and Variances of Early-Warning Characteristic Parameters under Various Operating Conditions
工况(Crq/MW) Pz/MPa Hz/m Tout/℃ Tin/℃ Qz/(kg·s−1) Rcv Gre/kW Rd 99.105 均值 15.4732 0.1574 326.489 292.92 98.15 0.35 132.7804 188 方差 1.135×10−5 0.0014 0.0039 0.0066 2.26×10−6 0.003 22.8991 1.0×10−6 99.35 均值 15.4807 0.132 326.667 293.12 98.349 0.3315 135.322 188 方差 3.381×10−4 0.0008 0.0125 0.0147 0.1097 1.2×10−3 100.1 1.0×10−6 99.97 均值 15.4846 0.082 327.069 292.785 98.04 0.3411 126.26 188 方差 4.17×10−5 0.0017 0.0481 0.0096 0.0921 1.02×10−2 142 1.0×10−6
下载: 导出CSV
表 4 不同工况下预警特征参数的阈值
Table 4. Thresholds of Early-Warning Characteristic Parameters under Various Operating Conditions
工况(Crq) /MW Pz/MPa Hz/m Tout/℃ Tin/℃ Qz/(kg·s−1) Rcv Gre/kW Rd 99.105 上阈值 15.4903 0.1977 326.557 293.0176 98.1744 0.392 137.56 188 下阈值 15.4886 −0.1185 326.4811 292.8524 98.1236 0.298 127.99 188 99.35 上阈值 15.5261 0.1694 326.7862 293.2063 98.6623 0.366 145.32 188 下阈值 15.4614 0.1585 326.5918 292.9337 98.0146 0.296 125.31 188 99.97 上阈值 15.5167 0.1283 327.2003 292.8877 98.3828 0.44 138.20 188 下阈值 15.4753 0.0863 326.8334 292.6324 97.7172 0.241 114.31 188
下载: 导出CSV
表 5 特征参数的权重
Table 5. Weights of Characteristic Parameters
序号 特征参数 权重 1 Pz 0.1176 2 Hz 0.1101 3 Tout 0.1139 4 Tin 0.1271 5 Qz 0.1227 6 Rcv 0.2005 7 Gre 0.1209 8 Rd 0.0872
下载: 导出CSV
-
[1] CHEN X M, WU C X, WU Y, et al. Design and analysis for early warning of rotor UAV based on data-driven DBN[J]. Electronics, 2019, 8(11): 1350. doi: 10.3390/electronics8111350 [2] 向健平, 凌永志, 詹俊, 等. 基于SCADA系统的风电机组主轴承故障预警方法[J]. 电力科学与技术学报,2019, 34(3): 223-228. doi: 10.3969/j.issn.1673-9140.2019.03.028 [3] CHEN G S, ZHENG Q Z. Online chatter detection of the end milling based on wavelet packet transform and support vector machine recursive feature elimination[J]. International Journal of Advanced Manufacturing Technology, 2017, 95(5): 1-10. [4] JIAN Y F, QING X G, ZHAO Y, et al. Application of model-based deep learning algorithm in fault diagnosis of coal mills[J/OL]. Mathematical Problems in Engineering, 2020. https://doi.org/ 10.1155/2020/3753274 DOI: 10.1155/2020/3753274. [5] ZHOU F, XU P C, LIN K. Early warning analysis of online vibration fault characteristics of motor base screw loosening based on similarity measurement theory[J/OL]. Archive of Applied Mechanics, 2020. https://doi.org/ 10.1007/s00419-020-01820-1. DOI: 10.1007/s00419-020-01820-1. [6] 刘帅,刘长良,甄成刚. 基于数据分类重建的风电机组故障预警方法[J]. 仪器仪表学报,2019, 40(8): 1-11. [7] JIANG D, GONG J, GARE A. Design of early warning model based on time series data for production safety[J]. Measurement, 2017(101): 62-71. [8] 刘耘彰. 电厂风机故障诊断与预警研究[D]. 杭州: 浙江大学, 2019. [9] 牛玉广,李晓彬,张佳辉. 基于多元状态估计与自适应闽值的电站辅机故障预警[J]. 动力工程学报,2019, 39(9): 717-724. doi: 10.3969/j.issn.1674-7607.2019.09.005 [10] 潘瑾宜,杨婷,钱虹. 小型模块化反应堆冷却剂平均温度的预测控制方法[J]. 核动力工程,2020, 41(3): 62-67. [11] 陈娟,马栋梁,周涛,等. 基于灰色关联度的超临界水自然循环换热系数影响因素分析[J]. 核动力工程,2017, 38(2): 19-23. [12] 张艳艳. 基于熵权与灰色关联融合分析的模拟电路故障诊断研究[D]. 北京: 北京交通大学, 2019. [13] 叶林,张亚丽,巨云涛,等. 用于含风电场的电力系统概率潮流计算的高斯混合模型[J]. 中国电机工程学报,2017, 37(15): 4379-4387+4578. -
计量
- 文章访问数: 168
- HTML全文浏览量: 99
- PDF下载量: 24
- 被引次数: 0
