不完备检测条件下的系统测试建模与分析

System Test Modeling and Analysis under Imperfect Detection

复杂装备系统故障在实际中无法进行完全检测和故障隔离,传统的基于完备检测假设的维修过程模型已不适用于系统性能分析。从实际出发建立了劣化系统在不完备检测条件下的测试过程Markov模型。根据Markov状态微分方程和系统稳态平衡理论,提出求解系统稳态可用度的一类解析算法,构建了系统功能故障前平均时间和平均维修费用计算模型。实例说明了系统的故障检测频率、故障检测率和虚警率等测试性参数对系统稳态可用度、功能故障前平均时间和平均维修费用的影响。

The traditional perfect--detection maintenance process models are unsuitable to analyze complex system performance, because the faults can＇t be detected and isolated perfectly in complex systems. A system testing process Markov model under the imperfect detection was developed for the deteriorating system. Based on Markov state differential equation and steady--state balance theory, an analytic algorithm was presented to calculate steady--state availability. A MTTFF and a MMC model were proposed. Influence of inspection rate, fault detection rate and fault alarm rate on the system steady--state availability, MTTFF and MMC was illustrated. The model can be used to determine testabiltiy index as a theoretical foundation.