当前位置：首页 > 过刊浏览->2025年第51卷第6期
基于多元线性回归的管输成品油质量衰减预测方法研究
Prediction method for quality attenuation of refined oil based on multiple linear regression
浏览(449) 下载(5)
- DOI:
- 10.3969/j.issn.1001-2206.2025.06.007
- 作者:
- 江璐鑫, 王玉彬, 闫锋, 李政兵, 陈思杭, 莫琳琳
JIANG Luxin, WANG Yubin, YAN Feng, LI Zhengbing, CHEN Sihang, MO Linlin
- 作者单位:
- 国家石油天然气管网集团有限公司科学技术研究总院分公司, 天津 300457
PipeChina Institute of Science and Technology, Tianjin 300457, China
- 关键词:
- 成品油管道; 顺序输送; 质量控制; 多元线性回归; 质量跟踪
refined oil pipeline;sequential transportation;quality control;multiple linear regression;quality tracking
- 摘要:
- 成品油在管道顺序输送过程中，后行油品质量会随着输送距离的增加而降低，存在关键质量指标（柴油闪点、汽油终馏点）衰减导致油品质量不合格的风险。针对顺序输送过程中存在的油品关键质量指标衰减问题，以国内典型成品油管道为例，通过油品质量跟踪实验、多元线性回归方法，对管输油品关键质量指标衰减规律开展研究，建立了管输0#柴油闪点、92#汽油终馏点的衰减预测模型，并验证了模型预测的准确性。结果表明，所建立的预测模型对管输柴油闪点、汽油终馏点衰减值的预测准确度均较高。其中，预测的柴油闪点衰减值平均偏差为0.39℃，最大偏差不超过±1.04℃；汽油终馏点衰减值平均偏差为0.54℃，最大偏差不超过±1.38℃。该模型可为制订油品进线质量标准、保障成品油管道经济且高质量运行提供参考。
During sequential transportation of refined oil in pipelines, the quality of oil products decreases with the increase in transportation distance, risking unqualified oil products caused by the attenuation of key quality indicators. In response to the attenuation of key oil quality indicators（diesel flash point and gasoline final boiling point） in the process of sequential transmission, with typical domestic refined oil pipeline as an example, a study was carried out on the attenuation law of key indicators of pipeline oil through oil quality tracking experiments and multivariate linear regression. A prediction model was built for the attenuation of diesel flash point and gasoline final distillation point in the process of pipeline transportation and the accuracy of the model prediction was verified. The results show that the established prediction model is highly accurate in predicting the attenuation values of flash point and final boiling point in the pipeline transmission process. Among them, the average deviation of flash point attenuation value prediction is 0.39 ℃, with the maximum deviation not exceeding ±1.04 ℃. The average deviation of the predicted attenuation value of the final distillation point is 0.54 ℃, with the maximum deviation not exceeding ±1.38 ℃. The model can provide a reference for formulating the quality standard of oil product inlet and guaranteeing the economic and efficient operation of oil product pipelines.