基于人工神经网络的石蜡熔点预测  

Prediction of Paraffin Melting Point Based on Artificial Neural Network

作　　者：纪博睿 赵晓弘 胡西娅 潘英妮[4] 王硕 JI Borui;ZHAO Xiaohong;HU Xiya;PAN Yingni;WANG Shuo(Liaoning Inspection,Examination&Certification Centre,Shenyang Liaoning 110032,China;Liaoning Product Quality Supervision and Inspection Institute,Shenyang Liaoning 110032,China;Key Laboratory of Testing and Quality Control for Petroleum Products,State Administration for Market Regulation,Shenyang Liaoning 110144,China;School of Traditional Chinese Medicine,Shenyang Pharmaceutical University,Shenyang Liaoning 110016,China)

机构地区：[1]辽宁省检验检测认证中心,辽宁沈阳110032 [2]辽宁省产品质量监督检验院,辽宁沈阳110032 [3]国家市场监督管理总局重点实验室(石油产品检测与质量控制),辽宁沈阳110144 [4]沈阳药科大学中药学院,辽宁沈阳110016

出　　处：《当代化工》2025年第1期202-205,共4页Contemporary Chemical Industry

基　　金：辽宁省自然科学联合基金项目(项目编号:2022-NLTS-12-03);2022年辽宁省市场局科学计划项目(项目编号:2022JS003);国家市场监管总局科技计划项目(项目编号:2023YJ14);2024年辽宁省市场局科学计划项目(项目编号:2024JS008)。

摘　　要：对于特种石蜡产品来说,有些石油蜡性能的微小差异就会显著影响产品的性能。为了缩短特种石蜡研发时间,提高产品开发效率,更好地指导特种石蜡的生产,提出了基于人工神经网络的石蜡熔点预测模型。以石蜡化学组成C_(18)~C_(37)、异构烷烃、环烷烃、烯烃的质量分数为输入,以总熔点为输出,建立了石蜡熔点神经网络模型。得到的熔点神经网络模型对石蜡熔点预测准确率为99.2646%,表明熔点神经网络模型能精确预测石蜡熔点。For special paraffin products,some slight differences in the performance of petroleum waxes will significantly affect the performance of the products.In order to shorten the research and development time of special paraffin,improve the product development efficiency and better guide the production of special paraffin,a prediction model of paraffin melting point based on artificial neural network was proposed.Taking the chemical composition of paraffin,C_(18)~C_(37),isoparaffins,cycloalkanes and olefins as input and the total melting point as output,a neural network model of paraffin melting point was established.The prediction accuracy of melting point neural network model for paraffin wax is 99.2646%,which shows that the melting point neural network model can accurately predict the melting point of paraffin wax.

关 键 词：人工神经网络 石蜡 熔点 算法 计算机模拟 预测 

分 类 号：TQ9[化学工程] TP183[自动化与计算机技术—控制理论与控制工程]