间歇状态监测下缓慢退化系统的剩余寿命预测与维修策略优化研究

发布时间：2018-03-09 07:00

本文选题：退化系统　切入点：Wiener过程　出处：《北京交通大学》2017年硕士论文　论文类型：学位论文

【摘要】：设备在长期运行过程中,受磨损、疲劳、腐蚀等影响,零部件丧失最初规定的功能,其性能及健康状态不可避免的发生退化,最终导致设备失效。此类具有缓慢失效现象的系统称之为缓慢退化系统。在缓慢退化过程中,设备质量下降、运维成本提高;零部件失效会增加事故发生概率,造成经济损失、人员伤亡等不可估量的后果。因此,在系统运行过程中,监测退化数据并实施预测与健康管理,及时对退化零部件采取有效的维修措施,对于切实保障退化系统的运行安全性、可靠性与经济性具有重要意义。鉴于此,本文以缓慢退化系统为研究对象,研究基于间歇状态监测的剩余寿命预测方法及控制限维修策略,主要进行了以下研究工作:首先,获取的退化数据受测量误差的影响,因此,选择卡尔曼滤波去除间歇监测数据的测量误差,通过实例验证该方法用于去除缓慢退化系统测量误差的可行性。其次,建立Wiener过程退化模型,并将连续Wiener过程退化模型转化为离散形式,然后提出一种基于Wiener过程与离散卡尔曼滤波的混合退化模型的剩余寿命预测方法。以某编组闸片间歇状态监测数据为例验证上述方法的可行性,为后续缓慢退化系统的维修策略研究奠定基础。再次,在SMDP的框架下,将闸片的退化过程进行离散化处理,转化成有限的退化状态,以单位期望成本最低为目标构建预防性控制限维修优化模型,由Wiener过程的性质确定状态转移概率并采用策略迭代算法求解模型。最后,与传统的基于寿命的维修策略作对比研究,验证本文所提出的控制限维修策略的正确性与有效性,并针对预防性控制限维修策略对相关参数进行灵敏度分析。
[Abstract]:During the long term operation, the equipment is affected by wear, fatigue, corrosion and so on. The components lose their original function, and their performance and health state inevitably degrade. This kind of system with slow failure phenomenon is called slow degradation system. In the process of slow degradation, the quality of equipment decreases and the cost of operation and maintenance increases; the failure of parts increases the probability of accidents and causes economic losses. Therefore, monitoring degradation data and implementing prediction and health management, timely and effective maintenance measures for degraded parts and components can effectively guarantee the operational safety of degraded system. The reliability and economy are of great significance. In view of this, the residual life prediction method based on intermittent state monitoring and the control limit maintenance strategy are studied in this paper. The main research work is as follows: first of all, The acquired degraded data is affected by the measurement error. Therefore, Kalman filter is selected to remove the measurement error of intermittent monitoring data, and the feasibility of this method to remove the measurement error of slow degradation system is verified by an example. The degenerate model of Wiener process is established, and the degenerate model of continuous Wiener process is transformed into discrete form. Then a method for predicting the residual life of a hybrid degradation model based on Wiener process and discrete Kalman filter is proposed. The feasibility of the method is verified by taking the intermittent state monitoring data of a marshalling sluice as an example. It lays a foundation for the following research on the maintenance strategy of slow degradation system. Thirdly, under the framework of SMDP, the degradation process of the gate is discretized and transformed into a finite degradation state. A preventive control limited maintenance optimization model is constructed with the lowest expected cost per unit. The state transition probability is determined by the nature of the Wiener process and the model is solved by the strategy iterative algorithm. Compared with the traditional maintenance strategy based on service life, the correctness and effectiveness of the proposed control limit maintenance strategy are verified, and the sensitivity analysis of the related parameters is carried out for the preventive control limit maintenance strategy.
【学位授予单位】：北京交通大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TB114.3

【参考文献】