摘要
成品油易挥发、易黏附、易渗透等特点,导致其在运输、储存及销售各过程中均存在损耗的问题。为了提升成品油零售环节损耗管理的精准性与时效性,采用了统计过程控制理论(SPC)与博弈组合赋权法,构建了基于指数加权移动平均(EWMA)控制图的多因素监控模型;剖析了成品油在运输、储存及销售过程中的6类损耗因子(大呼吸损耗、移库倒库损耗、小呼吸损耗、付油损耗、残漏蒸发损耗及其他合理损耗);通过层次分析法(AHP)与因子分析法分别确定主客观权重,并基于博弈论优化组合权重。研究结果表明:①EWMA控制图对微小损耗偏移的检测灵敏度较传统修哈特控制图提升32%,异常识别响应时间缩短至24 h内;②组合赋权模型显示,付油损耗与大呼吸损耗为关键影响因素,其异常贡献度超过总损耗的62%;③实例验证中,模型在25组数据中精准预警3次异常,误报率低于5%;④通过标准化作业流程与员工培训,人为操作误差减少40%。结论认为,该模型实现了非线性高噪声环境下的成品油微弱损耗信号提取,为成品油企业低成本、高质量管理提供了数据驱动的决策支持。
Refined oil possesses inherent characteristics such as volatility,adhesiveness,and permeability,resulting in measurable losses throughout its transportation,storage,and distribution phases.To enhance the precision and timeliness of loss management within the refined oil retail sector,this study integrates Statistical Process Control(SPC)theory with game-theoretic combination weighting to construct a multi-factor monitoring model based on the Exponentially Weighted Moving Average(EWMA)control chart.Six categories of loss factors were systematically analyzed:working loss(during tank filling/emptying),transfer loss(during inventory relocation),standing loss(from vapor breathing),dispensing loss,residual/leakage/evaporation loss,and other justifiable losses.Subjective and objective weighting coefficients were determined via Analytic Hierarchy Process(AHP)and Factor Analysis,respectively,with subsequent optimization of composite weights using game theory.The research results show that:①The detection sensitivity of the EWMA control chart to small loss deviations is 32%higher than that of the traditional Shewhart control chart,and the abnormal identification response time is shortened to within 24 hours;②The combination weighting model shows that fuel dispensing loss and large breathing loss are the key influencing factors,with their abnormal contribution exceeding 62%of the total loss;③In the instance verification,the model accurately warned of 3 abnormal situations in 25 groups of data,with a false alarm rate of less than 5%;④Through standardized operation procedures and employee training,human operation errors were reduced by 40%.The conclusion is that this model achieves the extraction of weak loss signals in a nonlinear and high-noise environment,providing data-driven decision support for low-cost and high-quality management of refined oil enterprises.
作者
张涛
谭晶纯
尹子钊
ZHANG Tao;TAN Jingchun;YIN Zizhao(School of Economics Management,China University of Petroleum(East China),Qingdao,Shandong 266580,China;Daqing Iron Man College,Daqing,Heilongjiang 163458,China)
出处
《世界石油工业》
2025年第5期106-112,共7页
World Petroleum Industry
基金
国家自然科学基金项目“不对称信息下项目调度和资源规划的多方协同优化研究”(71801224)。
关键词
统计过程控制
成品油损耗
EWMA控制图
博弈组合赋权
零售管理
statistical process control
refined oil loss
EWMA control chart
game-theoretic combination weighting
retail management
