SIA OpenIR  > 数字工厂研究室
数据驱动的变频器健康状态评估与预测
Alternative TitleData Driven Health Assessment and Prediction of Inverter
闫会玉1,2
Department数字工厂研究室
Thesis Advisor宋宏
Keyword异常识别 设备健康管理 数据驱动 健康评估 健康预测
Pages63页
Degree Discipline计算机应用技术
Degree Name硕士
2019-05-17
Degree Grantor中国科学院沈阳自动化研究所
Place of Conferral沈阳
Abstract设备的健康状态评估与预测是设备故障预测与健康管理(PHM)技术的主要研究内容之一,是保障设备稳定、安全运行的主要技术手段。然而国内外的学者往往都在研究大型电力设备、航空航天等军用设备,鲜有人去研究变工业频器设备的健康问题。随着国家节能减排政策的出台,变频器作为工业用电节能的主要装置,市场潜力巨大,因此变频器生产商也越来越重视针对变频器设备的故障预测和健康管理方法,本文在这样的背景下,对变频器设备的健康状态评估与预测问题进行了研究。由于工业变频器设备长期稳定运行,设备数据量庞大,并且数据价值密度不高,为降低数据分析难度,提高数据价值密度,本文首先对其进行了压缩存储。然后在过去几十年设备健康、设备维护方面的研究基础上,利用大量的设备运行数据,通过密度峰值聚类的方法对变频器异常进行了识别。接着基于变频器健人工康评分数据与异常识别数据,利用回归拟合寻找出了日异常率与设备健康的关系。最后通过ARIMA模型对设备未来的健康走势进行了有效的预测。本文搭建了设备健康管理分析平台,平台已在变频器设备中进行了试运行,用户可以通过平台观察设备运行参数情况,也可以通过调用模型查看或预测设备的健康值,同时平台还提供了一些能耗、以及故障诊断方面的模型。目前平台运行稳定,能够为设备管理人员提供所需的帮助信息,提高设备健康管理水平,促进设备安全可靠运行,缩减设备维护成本,提高设备利用率。
Other AbstractAssessment and prediction of equipment health status is one of the main research contents of prognostic and health management (PHM) technology, which is the main technical means to ensure stable and safe operation of equipment. However, scholars at home and abroad are often studying large-scale power equipment, aerospace or other military equipment, and few people to study the health of industrial frequency converter equipment. With the promulgation of national energy saving and emission reduction policy, as the main device of industrial energy saving, the frequency converter has great market potential. Therefore, the manufacturers of frequency converter pay more attention to the prognostics and health management methods of frequency converter equipment. Under this background, this paper studies the health status evaluation and prediction of frequency converter equipment. Due to the long-term stable operation of industrial frequency converter equipment, the amount of equipment data is very huge and the value density of data is low. In order to reduce the difficulty of data analysis and improve value density of data, this paper compressed and stored the operational data firstly. Then, based on the research of equipment health and equipment maintenance in the past decades, using a large number of equipment operation data, the frequency converter anomalies are identified by density clustering method. Then, based on the manual health scoring data and anomaly recognition data of the frequency converter, the relationship between daily anomaly rate and equipment health is found by regression fitting. Finally, the ARIMA model is used to predict the future health trend of equipment effectively. This paper builds an analysis platform for equipment health management. The platform has been tested in the frequency converter equipment. Users can observe the operation parameters of the equipment through the platform, and they can also view or predict the health value of the equipment by calling the model. At the same time, the platform also provides some models of energy consumption and fault diagnosis. At present, the platform runs steadily, which can provide the necessary help information for equipment managers, improve the level of equipment health management, promote the safe and reliable operation of equipment, reduce the cost of equipment maintenance and improve the utilization rate of equipment.
Language中文
Contribution Rank1
Document Type学位论文
Identifierhttp://ir.sia.cn/handle/173321/25166
Collection数字工厂研究室
Affiliation1.中国科学院沈阳自动化研究所
2.中国科学院大学
Recommended Citation
GB/T 7714
闫会玉. 数据驱动的变频器健康状态评估与预测[D]. 沈阳. 中国科学院沈阳自动化研究所,2019.
Files in This Item:
File Name/Size DocType Version Access License
数据驱动的变频器健康状态评估与预测.pd(1814KB)学位论文 开放获取CC BY-NC-SAApplication Full Text
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[闫会玉]'s Articles
Baidu academic
Similar articles in Baidu academic
[闫会玉]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[闫会玉]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.
 

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.