Web近红外光谱原油快评技术开发与应用  

Development and Application of Web-Based Near Infrared Crude Oil Fast-Evaluation Technology

作　　者：李敬岩[1] 褚小立[1] 刘丹[1] 董力军[2] 冯保杰[3] 许育鹏[1] 陈瀑[1] LI Jingyan;CHU Xiaoli;LIU Dan;DONG Lijun;FENG Baojie;XU Yupeng;CHEN Pu(Research Institute of Petroleum Processing,SINOPEC,Beijing 100083,China;Luoyang Petrochemical Company,SINOPEC,Luoyang 471000,China;Tianjin Petrochemical Company,SINOPEC,Tianjin 300000,China)

机构地区：[1]中国石化石油化工科学研究院,北京100083 [2]中国石化洛阳分公司,河南洛阳471000 [3]中国石化天津分公司,天津300000

出　　处：《石油学报（石油加工）》2022年第3期710-717,共8页Acta Petrolei Sinica(Petroleum Processing Section)

摘　　要：开发了Web版原油快评软件,在软件构架层面,由原油快评系统用户应用模块、原油快评近红外光谱库管理模块、原油快评建模模块、原油快评模型库管理模块和用户管理模块构成。该软件实现了模型库的统一维护和统一发布、光谱数据库和评价数据库的统一管理,对现有的原油快评算法进行了升级,经过对升级算法的综合调试使用,可明显提高预测结果的计算速度和识别准确率。收录石化企业A、B常加工原油并分别建立了石化企业A、B原油近红外光谱数据库,将石化企业原油近红外光谱数据库与中国石化石油化工科学研究院原油近红外光谱数据库进行合并,建立快评分析模型,对常加工油种进行监测分析,结果表明近红外光谱快评结果较为准确。Web-based crude oil fast-evaluation software was developed.At the software architecture level,it is composed of modules of user application,near infrared(NIR)spectrum library management,crude oil modeling,crude oil fast-evaluation model,library management and user management.The software has centralized system for maintenance and release of model database,management of spectral database and evaluation database,and upgrading the algorithm of crude oil fast evaluation.Through comprehensive application of the upgraded algorithm,calculation speed and recognition accuracy of prediction results can be significantly improved.NIR spectroscopy database of two refineries were built through collecting their regular processed crude oil’s NIR spectra.Then,the above database and Research Institute of Petroleum Processing(RIPP)database were combined to establish the fast evaluation analysis model to monitor and analyze crude oils in different refineries.It was found that NIR validation results are reliable.

关 键 词：近红外光谱 原油 网络 数据库 

分 类 号：O657.33[理学—分析化学]