摘要
针对全寿命周期设备的视情维护目标,建立了基于在线监测和离线分析的二级监测数据管理以及分析系统。该系统包含在线监测数据管理模块和离线分析数据管理模块。在线监测数据管理模块通过网络将在线监测数据存储在本地数据库,实现了在线数据的自动化管理,采用GM(1,1)等维新息预测模型实现了对在线油液指标数据发展趋势的实时预测,提取在线特征指标建立了支持向量机状态识别模型,实现了磨损状态的异常判断并给出离线分析建议。离线数据管理模块实现油液离线分析数据(理化、光谱、红外、分析铁谱、直读铁谱)的录入和维护,通过参考油液标准库评判油品性能的好坏,最终结合在线监测和离线分析结果做出视情维护决策的建议。
A two-stage data administration and intelligent analysis system, composed of on-line monitoring and off-line analysis modules, was established for equipment group life-cycle condition-based maintenance. On-line monitoring module was designed to store on-line data by a local database via internet and administrate the data automatically. A real-time trend was predicated from the on-line data using a GM （ 1,1 ） model. An intelligent wear state classification model was established using Support Vector Machine （SVM）. On-line character indexes were input into this model to diagnose the abnormal and give an off-line analysis advice. The off-line data administration module was designed as an interface of input and maintenance of off-line analysis indexes （including physiochemical, spectrum, analytical and directing reading ferro- graph indexes）, and the oil quality can be evaluated by referring to a criterion. A final decision of conditional maintenance based monitoring can be made by integrating the results of on-line monitoring and off-line analysis.
出处
《润滑与密封》
CAS
CSCD
北大核心
2009年第7期97-101,共5页
Lubrication Engineering
关键词
油液监测
数据管理
趋势分析
oil monitoring
data management
trend analysis
