多阶段马尔科夫模型及其在SIS安全评估中的应用  

Multistage Markov model and its application in safety assessment of SIS

作　　者：刘进东[1] 范铠[1] 涂煊[1,2] 

机构地区：[1]华东理工大学化工过程先进控制和优化技术教育部重点实验室,上海200237 [2]上海工业自动化仪表研究院工业过程自动化国家工程研究中心,上海200233

出　　处：《中国安全科学学报》2016年第12期80-85,共6页China Safety Science Journal

基　　金：上海市重点学科建设基金资助(B504);国家高技术产业发展项目(发改办高技[2013]2519号)

摘　　要：为综合分析检验测试因素对安全仪表系统(SIS)安全评估的影响,基于多阶段马尔科夫模型,提出一种SIS安全评估方法。首先,基于多指标优化方法确定检验测试策略,并构建运行期的马尔科夫状态转移模型以及涉及测试自身引起失效概率和非理想检验测试率的检验测试模型;然后,融合马尔科夫状态转移模型和检验测试模型得到多阶段马尔科夫模型;最后,将该模型应用于某循环流化床(CFB)锅炉压力保护系统的安全等级评估中。结果表明,多阶段马尔科夫模型能够综合考虑检验测试自身引起失效概率和非理想检验测试率对SIS失效率的影响;检验测试因素对SIS的安全性有较大影响,因此不容忽略。In order to comprehensively analyze the effect of proof test factors on safety assessment of SIS, the paper was aimed at developing a new method for assessing safety level of SIS. Firstly, the proof test strategy was determined based on the multi-induce optimization method. A Markov state transition model was built for the operation period and a proof test model was built. Failure probability due to test and rate of imperfect proof test were considered in the proof test model. Then, a multistage Markov model was built by combing the Markov model with the proof test model. Finally, the multistage Markov model was applied to assess safety level of the pressure protection system of a certain circulating fluidized bed （CFB） boiler. The results show that multistage Markov model can consider effects of proof test factors on the results safety assessment of SIS, and that safety of SIS is affected greatly by the proof test, so they should be ignored.

关 键 词：安全仪表系统(SIS) 安全完整性等级(SIL) 安全评估 马尔科夫模型 检验测试 

分 类 号：X913.4[环境科学与工程—安全科学] TB114.3[理学—概率论与数理统计]