论文部分内容阅读
基于事件树-动态故障树(ET-DFT)模型的动态概率安全评价(IDPSA)方法将ET和DFT结合,能充分描述系统动态行为对其可靠性与风险性的影响,但将DFT转化为马尔科夫链时,存在状态空间爆炸和易出错等问题。为此,建立基于离散时间贝叶斯网络(DTBN)的复杂系统DPSA方法,将ET-DFT模型转化为DTBN,给出各静态和动态逻辑门向DTBN转化的方法以及各逻辑门条件概率表(CPT)的计算方法。数控机床液压系统应用实例的分析验证结果表明,基于DTBN的DPSA方法既能得到系统故障时各个底事件发生故障的后验概率,又能得到系统薄弱环节在各时间区间内发生故障的概率,从而实现对系统较为精确的DPSA。
The dynamic probabilistic safety assessment (IDPSA) method based on Event Tree-Dynamic Fault Tree (ET-DFT) model combines ET and DFT to fully describe the influence of system dynamic behavior on its reliability and risk. However, Cove chain, there is state space explosion and error-prone and other issues. To this end, a DPSA method based on discrete time Bayesian Networks (DTBN) is established, and the ET-DFT model is transformed into DTBN. The method of transforming each static and dynamic logic gate into DTBN and the conditional probability table of each logic gate are given CPT) calculation method. The analysis and verification of numerical control machine tool hydraulic system shows that the DPSA method based on DTBN can not only get the posterior probability of failure of each bottom event when the system fails, but also get the probability of failure of system weakness in each time interval, DPSA to achieve more accurate system.