| 利用量子化学特征的模糊人工神经网络预测咪唑啉衍生物缓蚀效率 |
| |
| 引用本文： | 范峥,刘钊,井晓燕,姬盼盼,赵辉,康建.利用量子化学特征的模糊人工神经网络预测咪唑啉衍生物缓蚀效率[J].化工进展,2019,38(4):1961-1969. |
| |
| 作者姓名： | 范峥 刘钊 井晓燕 姬盼盼 赵辉 康建 |
| |
| 作者单位： | 西安石油大学化学化工学院,陕西西安,710065;中国石油抚顺石化公司石油三厂,辽宁抚顺,113001 |
| |
| 基金项目： | 陕西省科学技术研究与发展计划(2016GY-150);西安石油大学研究生创新与实践能力培养项目(YCS17211020) |
| |
| 摘 要： | 针对咪唑啉衍生物的量子化学特征参数与缓蚀效率存在复杂非线性关系，在利用多因素方差分析判断其相关性的基础上建立以最高占据轨道能量、最低未占据轨道能量、分子偶极矩、单点能、硬度、软度、亲核进攻指数、亲电进攻指数、电子转移参数以及咪唑环上非氢原子静电荷之和等量子化学特征参数为输入，以缓蚀效率为输出的模糊人工神经网络。结果表明，咪唑啉衍生物的量子化学特征参数及其缓蚀效率之间具有非常显著的相关性，据此所创建的Takagi-Sugeno型模糊人工神经网络预测模型采用10-30-1网络结构，通过Momentum优化算法对其进行反复训练直至其均方误差小于容许收敛误差限0.005，训练、测试阶段模型输出值与期望值近似呈线性关系，决定系数为0.9999，关联度较高，验证阶段该模型亦表现出良好的可靠性。因此利用量子化学特征的模糊人工神经网络预测模型能够准确预测不同咪唑啉系列衍生物的缓蚀效率。
|
| 关 键 词： | 咪唑啉衍生物 计算机模拟 量子化学 缓蚀效率 神经网络 |
| 收稿时间： | 2018-07-16 |
Prediction of corrosion inhibition efficiency of imidazoline derivatives using fuzzy artificial neural network based on quantum chemical characteristics |
| |
| Zheng FAN,Zhao LIU,Xiaoyan JING,Panpan JI,Hui ZHAO,Jian KANG.Prediction of corrosion inhibition efficiency of imidazoline derivatives using fuzzy artificial neural network based on quantum chemical characteristics[J].Chemical Industry and Engineering Progress,2019,38(4):1961-1969. |
| |
| Authors: | Zheng FAN Zhao LIU Xiaoyan JING Panpan JI Hui ZHAO Jian KANG |
| |
| Affiliation: | 1. College of Chemistry & Chemical Engineering, Xi’an Shiyou University, Xi’an 710065, Shaanxi, China
2. Fushun Petrochemical Company No. 3 Refinery, CNPC, Fushun 113001, Liaoning, China |
| |
| Abstract: | In order to build the complicated nonlinear relationship between quantum chemical characteristics of imidazoline derivatives and corrosion inhibition efficiency, the fuzzy artificial neural network adopting quantum chemical characteristics, including the highest occupied molecular orbital energy, the lowest unoccupied orbital energy, molecular dipole moment, single point energy, hardness, softness, nucleophilic attack index, electrophilic attack index, electron transfer parameter and the sum of static charges of non-hydrogen atoms on the imidazole ring as inputs, corrosion inhibition efficiency as outputs, was established to determine their correlation based on multi-factor variance analysis. The results revealed that there was a very significant correlation between the mentioned quantum chemical characteristics and the corrosion inhibition efficiency. With the help of above research, the obtained prediction model of Takagi-Sugeno fuzzy artificial neural network with 10-30-1 structure using momentum optimization algorithm was trained repeatedly until its mean square error less than convergence tolerance 0.005 was reached. The model output values were approximately linear with actual desired values in the training and testing stage and demonstrated superior correlation due to determination coefficient 0.9999. The good reliability of prediction model was also displayed in the validating stage. Therefore, the fuzzy artificial neural network model based on quantum chemical characteristics accurately predicted the capacities of corrosion inhibition efficiency of various imidazoline derivatives. |
| |
| Keywords: | imidazoline derivatives computer simulation quantum chemistry corrosion inhibition efficiency neural networks |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
| 点击此处可从《化工进展》浏览原始摘要信息 |
|
点击此处可从《化工进展》下载全文 |
|
