针对船用凝给水系统设备之间耦合关系较强,对该系统的研究只是选取部分参数而并非像设备一样基本涵盖全部特征参数,且该系统在实际运行过程中可以通过自调节来掩盖某些已发生的故障从而无法准确形成运行参数和故障间的映射关系这一现状,以传统单一机器学习算法为基础,通过拓展建立针对Stacking算法的多分类器性能评价指标,准确寻找运行参数和故障之间的映射关系,解决了多分类器性能评价难题。并利用样本数据设计出比较Stacking算法和单一算法综合性能的试验方法,验证了Stacking模型在凝给水系统故障诊断任务中的适用性和优越性。
In view of the strong coupling relationship between the equipment of Marine condensate feed water system, only some parameters were selected in the study of the system, instead of covering all characteristics like the equipment, and the status quo that the Condensate feed water system can cover up some existing faults through self-regulation in actual operation, so that the mapping relationship between operating parameters and faults cannot be accurately formed. Based on the traditional single machine learning algorithm, the performance evaluation indexes of multi-classifiers targeting at the stacking algorithm are expanded and established, and the mapping relationship between operating parameters and faults is accurately found. The multi-classification performance evaluation problem is solved. A test method that compares the comprehensive performance of stacking algorithm and a single stacking algorithm is designed, and the applicability and superiority of stacking model in fault diagnosis of condensate feed water system are verified.
2025,47(1): 138-142 收稿日期:2024-3-6
DOI:10.3404/j.issn.1672-7649.2025.01.024
分类号:U664.81
作者简介:陈砚桥(1978-),男,博士,副教授,研究方向为动力及热力系统的科学管理
参考文献:
[1] 李成钿, 何欧, 潘隆轩, 等. 凝给水系统故障分析及其诊断技术研究[J]. 化工自动化及仪表, 2022, 49(3): 334-338.
LI C D, HE O, PAN L X, et al. Condensate feed water system fault analysis and diagnosis technology research[J]. Journal of Chemical Industry Automation and Instrumentation, 2022, 49(3): 334-338.
[2] 肖林博, 陈辉, 管聪. 基于CAWOA-BP的船舶凝给水系统故障诊断[J]. 舰船科学技术, 2023, 45(6): 118-124.
XIAO L B, CHEN H, GUAN C. Fault diagnosis of marine condensate feed water system based on CAWOA-BP[J]. Ship Science and Technology, 2023, 45(6): 118-124.
[3] 王晓东, 马旭颖. 基于SVM模型算法和大数据分析技术的船舶设备故障诊断[J]. 上海船舶运输科学研究所学报, 2021, 44(1): 49-53.
WANG X D, MA X Y. Marine equipment fault diagnosis based on SVM model algorithm and big data analysis technology [J]. Journal of Shanghai Marine Transport Research Institute, 201, 44(1): 49-53.
[4] 宾世杨, 张振, 唐俊杰, 等. 基于机器学习的风电机组机械传动系统故障诊断研究[J]. 机械与电子, 2024, 42(1): 11-15.
BIN X Y, ZHANG Z, TANG J J, et al. Research on fault diagnosis of mechanical transmission system of wind turbine based on machine learning [J]. Machinery & Electronics, 2019, 42(1): 11-15.
[5] 赵瑶瑶. 基于机器学习的故障诊断方法研究综述[J]. 中国设备工程, 2024(1): 173-175.
ZHAO Y Y. Review of fault diagnosis methods based on machine learning[J]. China Equipment Engineering, 2024(1): 173-175.
[6] 王宏, 袁伯阳, 韩晨, 等. 基于机器学习的冷水机组早期故障诊断[J]. 低温与超导, 2023, 51(11): 96-102.
WANG H, YUAN B Y, HAN C, et al. Early fault diagnosis of chiller based on machine learning[J]. Cryogenics and Superconductivity, 2023, 51(11): 96-102.
[7] 李少丽. 基于机器学习的风力机叶片结冰故障诊断模型研究[D]. 西安: 西安理工大学, 2023.
[8] 汪凡雨, 吴一纯, 卜扬, 等. 基于机器学习的核电厂DCS卡件故障诊断研究[J]. 自动化仪表, 2023, 44(6): 5-12.
WANG F Y, WU Y C, BU Y, et al. Based on machine learning study on fault diagnosis of nuclear power plant DCS card[J]. Automation Instrument, 2023, 44(6): 5-12.
[9] 孙彤, 陈砚桥. 基于AHP的Stacking算法基分类器选择[J]. 兵工自动化, 2022, 41(1): 39-42.
SUN T, CHEN Y Q. Selection of stacking algorithm base classifiers based on AHP [J]. Automation of Ordnance Industry, 2002, 41(1): 39-42.
[10] 陈威, 谢成心, 侯冀超, 等. 集成学习经典算法研究[J]. 河北建筑工程学院学报, 2023, 41(3): 211-216.
CHEN W, XIE C X, HOU J C, et al. Research on classical algorithm of ensemble Learning[J]. Journal of Hebei University of Civil Engineering and Architecture, 2023, 41(3): 211-216.