基于二次指数平滑法的管道剩余壁厚预测

详细信息    查看全文 | 推荐本文 |

英文篇名：Prediction of Pipeline's Residual Thickness Based on Double Exponential Smoothing Method 作者：贺肖 ; 肖超波 英文作者：HE Xiao;XIAO Chao-bo;Guangzhou Special Pressure Equipment Inspection and Research Institute; 关键词：二次指数平滑法 ; 剩余壁厚 ; 权重系数 ; 方差分析 英文关键词：double exponential smoothing method;;residual thickness;;weight coefficient;;variance analysis 中文刊名：GDGS 英文刊名：Pipeline Technique and Equipment 机构：广州特种承压设备检测研究院; 出版日期：2018-07-15 出版单位：管道技术与设备 年：2018 期：No.152 语种：中文; 页：GDGS201804006 页数：4 CN：04 ISSN：21-1312/TH 分类号：26-28+33

摘要

为判断管道的安全状况,对管道寿命做出合理预测,建立了基于二次指数平滑法的预测模型。首先对管道壁厚数据进行拟合,选择最合适的平滑次数,再采用先大范围筛查再小范围求精的方法获得最优权重系数,进一步增加预测模型的准确性。对预测结果进行方差分析和误差判断,结果显示,权重系数取0.89时的均方误差为0.008 509 mm~2,对应2017年的预测误差为0.52%,预测结果能达到较高的精度。
        In order to judge the safety of the pipeline and to make a reasonable prediction of the pipeline life,a prediction model based on the double exponential smoothing method was established. First,fit the data of pipeline thickness,select the most appropriate smoothing times,and then use the method of first screening and small scale refinement to get the best weight coefficient,further increasing the accuracy of prediction model. The variance analysis and error judgement of the prediction results show that the mean square error is only 0. 008 509 mm2 when the weight coefficient is 0. 89. The prediction error corresponding to 2017 is only 0. 52%,and the prediction result can reach the high accuracy requirement.

引文

[1]何连,刘贤玉,徐婧,等.均匀腐蚀速率下内腐蚀管道最小剩余壁厚预测方法[J].油气储运,2017,36(5):508-512.
    [2]赵事,蒋晓斌,高惠临.腐蚀管道的失效概率和寿命预测方法[J].油气储运,2006,25(12):28-31.
    [3]王玉梅,郭书平.国外天然气管道事故分析[J].油气储运,2000,19(7):5-10.
    [4]帅健,张春娥,陈福来.腐蚀管道剩余强度评价方法的对比研究[J].天然气工业,2006(11):122-125.
    [5]LIU S Y,WILLIAM Q M.Statistical methods for estimating the minimum thickness along a pipeline[J].Technometrics,2015,57(2):164-179.
    [6]马娜,尹昌耕,秦金光.核电厂碳钢管道壁厚减薄原因分析[J].核动力工程,2012,33(6):123-125.
    [7]蔡克,瞿婷婷,马小芳,等.加载管线弯头壁厚减薄分析[J].管道技术与设备,2016(6):26-29.
    [8]乐钻,曹学文.内腐蚀管道最小剩余壁厚有限元分析[J].石油地质与工程,2012,26(3):131-133;136.
    [9]何俊,刘林兵,唐一举.基于线性二次指数平滑法的瓦斯含量预测[J].煤炭科学技术,2014,42(12):48-61.
    [10]王斌,王勤为,董科,等.基于二次指数平滑预测的虚拟机调度方法研究[J].计算机应用研究,2017,34(3):723-726.
    [11]刘真翔,陈晓慧.基于指数平滑和改进增量SVR的齿轮寿命预测研究[J].机械传动,2016,40(5):30-34.