文献类型：期刊文献

中文题名：自适应FOA-SVR在铝电解槽出铝量预测中的应用

英文题名：Application of Adaptive FOA-SVR in Aluminum Reduction Cell Output Prediction

作者：黎书文

第一作者：黎书文

机构：[1]贵州理工学院机械工程学院,贵州贵阳550003

第一机构：贵州理工学院机械工程学院

年份：2021

期号：7

起止页码：10-12

中文期刊名：机械设计与制造

外文期刊名：Machinery Design & Manufacture

收录：CSTPCD;;北大核心:【北大核心2020】；

基金：贵州省科技计划项目(黔科合平台人才[2017]5789-10)。

语种：中文

中文关键词：铝电解槽;产量预测;果蝇优化算法;支持向量回归机;自适应

外文关键词：Aluminum Electrolyzer;Yield Prediction;Fruit Fly Optimization Algorithm;Support Vector Regres?sion Machine;Adaptation

摘要：为进行铝电解槽出铝量预测,提出了一种基于自适应FOA-SVR模型的铝电解槽出铝量预测方法。首先,设计了一种自适应果蝇优化算法,基于三维搜索空间引入自适应步长,并应用于支持向量回归模型参数优化。其次,选取槽平均电压、槽电阻、槽温度、加料量、氟盐量等5个影响铝电解槽出铝量的因素作为模型输入向量,对电解槽出铝量进行预测。最后,以某公司铝电解槽生产数据为例进行实验验证,自适应FOA-SVR模型预测结果相比FOA-SVR模型提高了预测精度和收敛效率,为电解铝产量预测提供了有效模型,具有一定的推广价值。
In order to predict the aluminum output of aluminum reduction cells,an aluminum output prediction method based on the adaptive FOA-SVR model is proposed.Primarily,an adaptive Drosophila optimization algorithm was designed,which intro?duced an adaptive step size based on a three-dimensional search space and applied it to parameter optimization of the support vec?tor regression model.Secondly,five factors that affect the aluminum output of the aluminum electrolysis cell are selected as model input vectors to predict the aluminum output of the electrolytic cell,which include the average cell voltage,cell resistance,cell temperature,feed amount,and fluoride salt amount.Finally,taking the production data of a certain company′s aluminum elec?trolysis cell as an example for experimental verification,compared with the FOA-SVR model,the prediction result shows that the prediction accuracy and convergence efficiency are improved with the adaptive FOA-SVR model,which provides an effective model for the prediction of electrolytic aluminum output with certain promotion value.