基于Lasso-Wiener模型的埋地腐蚀管道可靠性评价

Reliability Evaluation of Buried Corroded Pipelines Based on Lasso-Wiener Model

为研究埋地腐蚀管道的可靠性,通过套索回归算法对影响管道腐蚀的关键因素进行数据提取和降维,建立基于维纳退化模型和加速应力方程的可靠性分析模型,并利用马尔科夫链-蒙特卡洛方法对未知参数进行估值。以某腐蚀管道为例,选取5个因素作为土壤腐蚀代表因素,管道可靠性先保持平稳后急剧减小,可靠度为90%的时长为12.5年,剩余寿命为14.7年,与实际情况相符。该模型是通过产品性能退化过程建立的可靠性模型,具有较好的适应性和准确性,研究结果为小样本数据的利用提供了实际参考。

In order to study the reliability of buried corroded pipelines,the Lasso Regression Algorithm is used to extract data and reduce dimension of key factors affecting pipeline corrosion.The reliability analysis model based on Wiener Degradation and Accelerated Stress Equation is established,and the unknown parameters are estimated by using the Markov Chain Monte Carlo method.Taking a corroded pipeline as an example,five factors are obtained as representative factors of soil corrosion,the reliability of the pipeline first stays stable and then decreases sharply.The time when the reliability reaches 90%is 12.5 years,and the remaining life is 14.7 years,which is consistent with the actual situation.The model establishes the reliability model through the product performance degradation process,which has good adaptability and accuracy.The research result provides a practical reference for the use of small sample data.