大连理工大学学报
大連理工大學學報
대련리공대학학보
JOURNAL OF DALIAN UNIVERSITY OF TECHNOLOGY
2013年
4期
479-484
,共6页
基于风险的检验(RBI)%风险%失效概率%不确定性量化%马尔可夫模型%灾害率%泄漏后破裂
基于風險的檢驗(RBI)%風險%失效概率%不確定性量化%馬爾可伕模型%災害率%洩漏後破裂
기우풍험적검험(RBI)%풍험%실효개솔%불학정성양화%마이가부모형%재해솔%설루후파렬
risk-based inspection (RBI)%risk%failure probability%uncertainty quantitative%Markov model%hazard rate%leakage before break (LBB)
一种基于泄漏后破裂(LBB )的马尔可夫模型被应用于研究过程工业中的维修活动对于管道失效概率的影响,该模型可以预测复杂维修方案(包含带压堵漏等内容)对于管道失效概率的影响,这对于维修方案的定量评估有重要意义。根据中国石化企业中的管道维修特点,将管道状态分为完好、可探测缺陷、泄漏和破裂4种状态,相应地建立了LBB四状态马尔可夫模型,进而推导出了在维修方案影响下管道失效概率随时间变化的常微分方程,并应用盛金公式求解该微分方程,最后通过一个管道减薄维修的案例验证了该马尔可夫模型计算结果的可靠性。
一種基于洩漏後破裂(LBB )的馬爾可伕模型被應用于研究過程工業中的維脩活動對于管道失效概率的影響,該模型可以預測複雜維脩方案(包含帶壓堵漏等內容)對于管道失效概率的影響,這對于維脩方案的定量評估有重要意義。根據中國石化企業中的管道維脩特點,將管道狀態分為完好、可探測缺陷、洩漏和破裂4種狀態,相應地建立瞭LBB四狀態馬爾可伕模型,進而推導齣瞭在維脩方案影響下管道失效概率隨時間變化的常微分方程,併應用盛金公式求解該微分方程,最後通過一箇管道減薄維脩的案例驗證瞭該馬爾可伕模型計算結果的可靠性。
일충기우설루후파렬(LBB )적마이가부모형피응용우연구과정공업중적유수활동대우관도실효개솔적영향,해모형가이예측복잡유수방안(포함대압도루등내용)대우관도실효개솔적영향,저대우유수방안적정량평고유중요의의。근거중국석화기업중적관도유수특점,장관도상태분위완호、가탐측결함、설루화파렬4충상태,상응지건립료LBB사상태마이가부모형,진이추도출료재유수방안영향하관도실효개솔수시간변화적상미분방정,병응용성금공식구해해미분방정,최후통과일개관도감박유수적안례험증료해마이가부모형계산결과적가고성。
A Markov model based on leakage before break (LBB) is applied to study the effects of maintenance activities on the failure probability of pipe systems in process industries .The model can predict the influence of complex maintenance plans (including the pressured seal ,etc .) on the pipe failure probability ,w hich is significant for quantitative assessment of maintenance plans .According to features of pipe maintenance in petrochemical industries in China , pipe is distinguished into four states , success state , flaw state , leak state and rupture state , similarly constant coefficient differential equations have been deduced on the pipe failure probability with the time after establishing a LBB four‐states Markov model ,and these constant coefficient differential equations can be solved with Shengjin formula .At last ,a practical case is calculated applying the model in a maintenance activity of pipe thinning to testify the reliability of the model .