基于马尔可夫链的埋地燃气钢管管壁腐蚀预测

Application of Markov chain in prediction of corrosion conditions for buried gas steel pipeline

摘要:

为适时对埋地燃气钢管腐蚀与防护系统现状进行科学评价,建立一种埋地燃气钢管管壁腐蚀状况预测模型.应用钢质管道管体腐蚀损伤评价方法和马尔可夫链理论,阐述利用转移概率矩阵对管壁腐蚀状况进行预测的方法.以某一埋地燃气管线为预测实例,计算获得该管线管壁腐蚀状态分布的预测值.结果显示,对该管壁腐蚀检测的最后腐蚀状态的计算值与实际检测值对应的腐蚀状态相吻合,同时预测结果也与拟合腐蚀深度常用的指数函数式预测的最大腐蚀深度计算值相一致.表明将马尔可夫链应用于管壁腐蚀状况预测是可行的,从而为合理确定管道检测和维修周期提供理论依据.

Abstract:

A prediction model of corrosion conditions is developed for buried gas steel pipelines in order to evaluate opportunely the existing states of their corrosion and prevention.A method of predicting corrosion states for the wall of buried gas pipelines by transition probability matrices is described by applying the evaluation method of corrosion damage of steel pipelines and the theory of Markov chain.Taking a buried gas pipeline for example,the corrosion states for the wall of the pipeline are predicted.It is shown that the calculated state for the latest detection of this pipe wall corrosion has a good agreement with corresponding state for its detected value and all predicted results also have an agreement with the corrosion depth calculated by the exponential function that commonly fits the max corrosion depth,which demonstrates that application of Markov chain in the prediction of corrosion is effective for buried pipelines.Thus it provides a theoretical basis for reasonable detection and maintenance period of buried gas pipelines.