Deeppipe: An intelligent framework for predicting mixed oil concentration in multi-product pipeline

被引:7
作者
Du, Jian [1 ]
Zheng, Jianqin [2 ]
Liang, Yongtu [1 ]
Xia, Yuheng [1 ]
Wang, Bohong [3 ]
Shao, Qi [4 ]
Liao, Qi [1 ]
Tu, Renfu [1 ]
Xu, Bin [1 ]
Xu, Ning [1 ,5 ]
机构
[1] China Univ Petr, Natl Engn Lab Pipeline Safety, MOE Key Lab Petr Engn, Beijing Key Lab Urban Oil & Gas Distribut Technol, Fuxue Rd 18, Beijing 102249, Peoples R China
[2] China Petr Planning & Engn Inst, Xinxi Rd 3, Beijing 100083, Peoples R China
[3] Zhejiang Ocean Univ, Natl & Local Joint Engn Res Ctr Harbor Oil & Gas S, Sch Petrochem Engn & Environm, Zhejiang Key Lab Petrochem Environm Pollut Control, 1 Haida South Rd, Zhoushan 316022, Peoples R China
[4] PipeChina South China Co, Linjiang Ave 1, Guangzhou 510623, Peoples R China
[5] China Univ Petr Beijing Karamay, Karamay 834000, Xinjiang, Peoples R China
关键词
Multi-product pipeline; Mixed oil concentration; Theory-guided feature engineering; Curve parameterization; Virtual samples generation; TREND-DIFFUSION; MILP MODEL;
D O I
10.1016/j.energy.2023.128810
中图分类号
O414.1 [热力学];
学科分类号
摘要
Accurately predicting mixed oil concentration distribution exerts a core effect on the optimization of pipelines and the quality of oil. Due to the neglect of mechanism features, high-dimensional complex feature correlations, and insufficient feature information on small batch data, the current methods cannot predict mixed oil concentration accurately. This work proposes a hybrid intelligent framework to provide an accurate and effective monitoring tool for mixed oil concentration of multi-product pipelines. In the proposed framework, the development mechanism of mixed oil is analyzed to select and reconstruct holistic features to explore the influencing mechanism of mixed oil concentration. Then, a parameterization and nonlinear transformation module is designed to acquire the accurate and concise representation of mixed oil concentration, thus decreasing the complexity of feature space and promoting the approximating ability of the prediction model. Eventually, a novel virtual samples generation module is established to obtain high-quality samples of mixed oil concentration, aiming to extract more comprehensive correlations of feature variables and improve the prediction performance. Cases from real-world multi-product pipelines suggest more accurate prediction results of mixed oil concentration compared to other advanced methods, with RMSE and R2 being 0.0500 and 0.9688. Furthermore, it is also proved that acquiring more holistic and accurate feature variables of mixed oil development and fully exploring comprehensive correlations between feature variables are crucial for the performance enhancement of the prediction model.
引用
收藏
页数:21
相关论文
共 52 条
[1]   Multi-period optimal schedule of a multi-product pipeline: A case study in Algeria [J].
Abdellaoui, Wassila ;
Souier, Mehdi ;
Sahnoun, M'hammed ;
Ben Abdelaziz, Fouad .
COMPUTERS & INDUSTRIAL ENGINEERING, 2021, 159
[2]  
Austin J.E., 1963, P I MECH ENG, V178, P377, DOI DOI 10.1177/002034836317800160
[3]  
Baptista RM, C MIX VOL EV PUMP SH
[4]  
Baptista RM, C EST MIX VOL BATCH, V40252
[5]   Application of a heterogenous multiscale method to multi-batch driven pipeline [J].
Blazic, Saso ;
Geiger, Gerhard ;
Matko, Drago .
APPLIED MATHEMATICAL MODELLING, 2014, 38 (03) :864-877
[6]   A new approach to prediction of radiotherapy of bladder cancer cells in small dataset analysis [J].
Chao, Gy-Yi ;
Tsai, Tung-I ;
Lu, Te-Jung ;
Hsu, Hung-Chang ;
Bao, Bo-Ying ;
Wu, Wan-Yu ;
Lin, Miao-Ting ;
Lu, Te-Ling .
EXPERT SYSTEMS WITH APPLICATIONS, 2011, 38 (07) :7963-7969
[7]   Energy Consumption Reduction and Sustainable Development for Oil & Gas Transport and Storage Engineering [J].
Chen, Xianlei ;
Wang, Manqi ;
Wang, Bin ;
Hao, Huadong ;
Shi, Haolei ;
Wu, Zenan ;
Chen, Junxue ;
Gai, Limei ;
Tao, Hengcong ;
Zhu, Baikang ;
Wang, Bohong .
ENERGIES, 2023, 16 (04)
[8]   Efficient subsampling of realistic images from GANs conditional on a class or a continuous variable [J].
Ding, Xin ;
Wang, Yongwei ;
Wang, Z. Jane ;
Welch, William J. .
NEUROCOMPUTING, 2023, 517 :188-200
[9]  
Du J, 2022, Chemical Eng. Trans., V94, P1165
[10]   Deeppipe: A two-stage physics-informed neural network for predicting mixed oil concentration distribution [J].
Du, Jian ;
Zheng, Jianqin ;
Liang, Yongtu ;
Xu, Ning ;
Klemes, Jiri Jaromir ;
Wang, Bohong ;
Liao, Qi ;
Varbanov, Petar Sabev ;
Shahzad, Khurram ;
Ali, Arshid Mahmood .
ENERGY, 2023, 276