摘要
针对设备群的故障预测与健康管理系统(PHM)技术的特点,提出了一种按设备级、工作单元级、设备群级和企业级分层的PHM功能模型,建立了功能模块间及不同层间的信息流动及需求关系。基于WEB服务和SOA技术,建立了PHM的系统分层架构模型;提出一种基于时间戳数据防过期方法,解决了基于XML的数据交互过程中的数据同步问题;研究了各功能组件模块之间的协同运行问题;基于逻辑斯蒂回归和支持向量机建立了性能退化趋势预测模型。在此基础上,开发了基于WEB的PHM软件原型系统,以验证上述方法和技术的可行性。
According to the characteristic of facilities group prognostics and health management systems, a layered PHM functional model is put forward in term of facility, workstation, facilities group and enterprise, the information relationship and requirements relationship are given at different layers or different function modules. Firstly, a layered PHM platform framework is established based on WEB Service and SOA technology. Secondly, based on time stamp technology, the data synchronization method is studied in an interoperation model based on XML. Thirdly, the collaborating problems between different functional component nodes are solved. Further more, a facilities degradation trend forecasting is established through logistic regression and support vector regression technology. Finally, a prototype of PHM software is developed to demonstrate the flexibility of above mentioned methods and technologies.
出处
《机械》
2008年第7期57-60,共4页
Machinery
基金
国家重点基础研究发展973计划资助(2005CB724106)
教育部新世纪优秀人才计划项目资助(NCET-04-0928)