不完备检测下的劣化系统视情维修建模与优化

Journal of Systems Engineering and Electronics ›› 2012, Vol. 34 ›› Issue (8): 1640-1645.

不完备检测下的劣化系统视情维修建模与优化

葛恩顺, 李庆民, 彭英武, 阮旻智

海军工程大学兵器工程系, 湖北武汉 430033

出版日期:2012-08-27 发布日期:2010-01-03

Condition based maintenance model and optimization under imperfect inspection for deteriorating systems

GE En-shun, LI Qing-min, PENG Ying-wu, RUAN Min-zhi

Department of Weaponry Engineering, Naval University of Engineering, Wuhan 430033, China

Online:2012-08-27 Published:2010-01-03

摘要/Abstract

摘要：

功能检测是对劣化系统开展视情维修的首要工作，由于技术条件和知识水平的限制，实际中对系统劣化状态的检测识别很难给出精确结果，势必影响视情维修工作的有效进行。针对这一问题，基于多级劣化系统检测维修马尔可夫链模型，通过对不完备检测因素影响下劣化系统寿命周期的3种情况进行分析，结合常见虚警和漏检两种检测错误，以长期运行费用率最低为目标，计算最优检测周期和维修阈值。通过实例，分析在检测不完备情况下检测周期和维修阈值对系统长期运行费用率的影响，并验证模型的有效性和合理性。

Abstract:

The functional inspection is of the first importance in the process of conditionbased maintenance for deteriorating systems. But it is hardly possible to acquire the system state accurately in reality, which is restricted by the technology and knowledge level. Aiming at this problem, based on the Markov chain model established for the multi-state deteriorating system, three situations of the deteriorating system life cycle under imperfect inspection are analyzed, and based on the functions of false alarm and the undetected error, the joint optimization of the inspection period and the threshold value of preventive maintenance are investigated for the minimization of the system long-run cost rate. Through a numerical example, the influence of the inspection period and threshold value on the system longrun cost rate under the imperfect inspection is analyzed. Therefore, the correctness and rationality of the model are verified.

葛恩顺, 李庆民, 彭英武, 阮旻智. 不完备检测下的劣化系统视情维修建模与优化[J]. Journal of Systems Engineering and Electronics, 2012, 34(8): 1640-1645.

GE En-shun, LI Qing-min, PENG Ying-wu, RUAN Min-zhi. Condition based maintenance model and optimization under imperfect inspection for deteriorating systems[J]. Journal of Systems Engineering and Electronics, 2012, 34(8): 1640-1645.