【作者】 江伟；

【导师】 罗毅；

【作者基本信息】 华中科技大学 ， 电力系统及其自动化， 2006， 硕士

【摘要】 电力系统的设备维修策略逐渐从以时间周期为基础的定期维修制度向以设备状态为基础的状态维修制度过渡,并取得了一定的成效。但由于对状态维修的定义及含义见仁见智,电气设备的状态维修在我国至今仍主要停留在学术讨论范围,且欠缺对状态维修策略的完整的探讨。本文在现有研究的基础上,着重对状态维修所涉及的关键技术展开系统的研究。论文首先回顾了国内外电气设备状态维修研究和应用的现状,分析了现有研究存在的一些误区及缺陷;将实施状态维修的关键技术概括为状态监测技术、状态预测技术及状态评估技术并阐述了它们的意义;理清了状态维修与在线监测及故障诊断的关系。论文分别从安全性和经济性的角度确定了状态维修所需的状态监测间隔期。采用时间序列分析方法预测电气设备状态,针对设备特征量序列的周期性和趋势性,建立了合适的ARIMA模型对该非平稳序列进行分析和预测。通过对某变压器油中气体含量的实例预测验证了模型的有效性。论文确定了指示设备状态的静态指标和动态指标,针对现有的电气设备状态评估方法存在的固有缺陷,提出了基于变权综合理论的新的状态评估方法,构造了强局部惩罚-激励型状态变权函数对某变压器的状态进行了实例评估。在此基础上,给出了电气设备状态维修决策的建议决策依据及对应的维修手段。论文还分别讨论了理想的及非理想的预防性维修对电气设备可靠性的影响,比较了定期维修和状态维修对设备可靠性影响的异同。最后对本文所做工作做了总结,指出文中研究的不足并对今后的研究方向进行了展望。更多还原

【Abstract】 The maintenance strategy of electrical equipments has gradually being changed from Time-based Maintenance (TBM) to Condition-based Maintenance (CBM). Although some effects have been achieved, the CBM strategy of electrical equipments is sticked in academic area in China because of the different understandings of the definition and signification of CBM. Moreover, some integrated discussions and researches of CBM strategy is wanting at the present time. Based on the existing researches, this paper does some systemic research on the key techniques of CBM chiefly.Firstly, this paper generalizes the academic and practical status of domestic and foreign Condition-based Maintenance strategy of electrical equipments. Some mistakes and bugs of present researches are analyzed. The key techniques of CBM are generalized to condition monitoring technique, condition forecasting technique and condition assessment technique. The function of each technique is expatiated, while the connections among CBM, online monitoring and fault diagnosis has been made clear.Security-based and economy-based methods for confirming the time interval of condition monitoring in CBM are given in this paper. Time Serial Analysis Method is used to predict future condition of electrical equipments. Appropriate ARIMA model is constructed to analysis and predict the characteristic parameter time serial which is periodic and trended. The validity of this ARIMA model is approved by predicting the gas-in-oil parameter of an electric transformer. Then, a new condition assessment method based on variable weights analysis is presented. An assessment example is given while a local variable weight function being constructed. Based on this result, some recommendable CBM decision-making values and maintenance operations are listed in this paper. Effects on the reliability of equipments by TBM and CBM strategy are discussed in both ideal and non-ideal cases.At the end of this dissertation, the main results of the research are summarized. And point out the deficiency existed in research process. The further work is submitted as well.更多还原