基于概率权重灰色马尔可夫模型的腐蚀预测
Corrosion Prediction Based on Probability Weighted Gray Markov Model
郑宇佳 1董增寿 1张晓红 2石慧1
作者信息
- 1. 太原科技大学电子信息工程学院,山西 太原 030024
- 2. 太原科技大学经济与管理学院,山西 太原 030024
- 折叠
摘要
精准地预测管道的腐蚀发展趋势是降低油气事故损失,保证其安全运输的主要途径之一.针对常用的基于聚类中心灰色马尔可夫预测模型中等转移概率会导致较大预测误差的问题,提出了一种基于概率权重的灰色马尔可夫模型用于油气管道腐蚀预测.首先用灰色模型对管道腐蚀状态进行初步预测,后采用模糊C均值聚类对预测误差数据聚类,并提出基于概率权重的马尔可夫模型对预测值进行修正,以提高出现等转移概率时模型的预测精度.最后通过两个实例验证了上述模型的有效性和正确性,并与基于聚类中心灰色马尔可夫模型进行比较.
Abstract
Accurately predicting the corrosion development trend of pipelines is one of the main ways to reduce the loss of leakage accidents and ensure their safe production and transportation.In this paper,a gray Markov model based on probability weights is proposed to predict the corrosion of oil and gas pipelines.Firstly,the gray model was used to make a preliminary prediction of the corrosion state of the pipeline.Then,the fuzzy C-means clustering was used to cluster the prediction error data,and a Markov model based on probability weight was proposed to improve the prediction accuracy in the case of equal transition probability.Finally,the effectiveness and correctness of the above model were verified through two examples,and compared with the gray Markov model based on the cluster center.
关键词
管道腐蚀/灰色马尔可夫/聚类/概率权重Key words
Pipeline corrosion/Gray Markov/Clustering/Probability weight引用本文复制引用
基金项目
国家自然科学基金(72071183)
国家自然科学基金(71701140)
国家自然科学基金(61703297)
山西省重点研发计划(201903D321012)
山西省重点研发计划(201903D121023)
山西省留学回国人员科技活动择优资助项目(2018002)
出版年
2024
NETL
NSTL
万方数据