以可靠性为中心的设备维修决策技术研究与应用

发布时间：2018-01-03 12:33

本文关键词：以可靠性为中心的设备维修决策技术研究与应用　出处：《北京化工大学》2012年硕士论文　论文类型：学位论文

【摘要】：现代企业中，典型的维修问题有被动的维修，重复的检修，合理的维修无法体制化，不必要和保守的预防性维修，以及缺乏可追溯性的维修计划等，因此，发展一种合适的维修方法来处理这类问题十分必要。以可靠性为中心的维修（RCM）提供了最系统和最有效的方式可实现工厂和设备维护计划的优化，帮助制定合理且有效的维修决策。目前，，RCM已初步在我国的多个炼化企业中得到应用，但由于各种原因，其实施应用的效果并不理想，主要表现在很难利用已知的历史数据或状态监测数据获得设备的隐含信息进行风险的评估，为此，本文从以下方面进行了研究：（1）将一元和混合威布尔寿命分布模型引入到RCM的风险评估中，研究相应的参数估计方法，总结模型的选用原则，利用所得结果分析可靠性的各指标参量，并对中石油往复压缩机组关键部件进行风险评估，取得了较理想的效果。（2）引进三次指数平滑、灰色马尔科夫和灰色关联分析的神经网络法，实现状态监测下小样本波动数据的故障分析，以某轴承的数据验证了方法的有效性和可行性，另外，也探索性地研究其剩余寿命的估算。（3）详细介绍了RCM对中石油往复压缩机关键零部件的实施过程，并将上述各模型应用到该评估中，根据评估结果筛选出了可能产生高中风险故障模式对应的关键零部件，同时完成维修任务的优化。
[Abstract]:In modern enterprises, the typical maintenance problems are passive maintenance, repeated maintenance, reasonable maintenance can not be institutionalized, unnecessary and conservative preventive maintenance, and lack of traceability maintenance plan, and so on. It is necessary to develop an appropriate maintenance method to deal with such problems. Reliability centered maintenance (RCM) provides the most systematic and efficient way to optimize plant and equipment maintenance plans. Help to make reasonable and effective maintenance decisions. At present, RCM has been initially applied in many refinery and chemical enterprises in China, but the effect of its application is not ideal because of various reasons. The main performance is that it is difficult to use known historical data or condition monitoring data to obtain the hidden information of the equipment for risk assessment. Therefore, this paper studies the following aspects: 1) the univariate and mixed Weibull life distribution model is introduced into the risk assessment of RCM, the corresponding parameter estimation method is studied, and the selection principle of the model is summarized. The reliability parameters are analyzed by using the obtained results, and the risk assessment of key components of CNPC reciprocating compressor unit is carried out, and the satisfactory results are obtained. 2) the neural network method of cubic exponential smoothing, grey Markov and grey correlation analysis is introduced to realize the fault analysis of small sample fluctuation data under state monitoring. The validity and feasibility of the method are verified by the data of a bearing. In addition, the estimation of its residual life is also studied in an exploratory way. The implementation process of key components of CNPC reciprocating compressor by RCM is introduced in detail, and the above models are applied to the evaluation. According to the evaluation results, the key parts corresponding to the risk failure mode of high and high school were selected, and the maintenance task was optimized at the same time.
【学位授予单位】：北京化工大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TH17

【参考文献】