摘要
针对调和汽油辛烷值建模中的变量选择问题、模型适应性问题与辛烷值的优化问题,采用随机森林、最大信息系数与皮尔森相关系数组合提出了一种辛烷值建模变量选择的方法。还提出一种基于BP神经网络与模糊神经网络的建模方法,建立对辛烷值的预测模型,提高了辛烷值预测模型的适应性。在此基础上,对基本粒子群算法进行了改进,改进后的粒子群算法更符合实际生产对操作变量允许调整幅度值为确定值的要求,并且提高了优化算法的计算速度。
Considering the variable selection in modeling the blended gasoline octane number,and the model adaptability and the optimization of the octane number,having the combination of random forest,maximum information coefficient and Pearson correlation coefficient used to propose an octane number modeling variable choosing method was implemented,including a modeling method based on BP neural network and fuzzy neural network to establish a prediction model for octane number so as to improve the adaptability of the octane number prediction model.On this basis,a modified particle swarm optimization(PSO)algorithm was proposed.The modified PSO algorithm can satisfy the actual production requirements that the allowable adjustment value of the operating variable has to be a certain value,and it improves the calculation speed of the optimization algorithm.
作者
谢忻南
饶伟浩
薛美盛
XIE Xin-nan;RAO Wei-hao;XUE Mei-sheng(School of Information Science and Technology,University of Science and Technology of China)
出处
《化工自动化及仪表》
CAS
2022年第1期60-67,共8页
Control and Instruments in Chemical Industry
