论文部分内容阅读
针对铝电解过程电流效率预测问题,建立了一种电流效率智能集成预测模型。首先,基于铝电解过程及其数据的特点,以及模糊c-均值聚类和非监督聚类方法的不足,提出一种模糊c-均值监督聚类改进算法对其聚类,并在此基础上建立了监督式分布支持向量机智能预测模型。其次,基于铝电解反应原理,建立了电流效率机理预测模型。最后,对两种模型进行加权集成,得到电流效率智能集成模型,并利用现场生产数据进行仿真验证,结果表明其预测精度较高,可用于铝电解电流效率实际生产预报。
Aiming at the current efficiency prediction of aluminum electrolysis process, an intelligent integrated prediction model of current efficiency is established. First of all, based on the characteristics of aluminum electrolysis process and its data, as well as the deficiencies of fuzzy c-means clustering and unsupervised clustering methods, a fuzzy c-means supervised clustering algorithm is proposed to improve clustering. Based on this, Established a supervised distributed support vector machine intelligent prediction model. Secondly, based on the principle of aluminum electrolysis, the current efficiency model is established. Finally, the two models are weighted integrated to obtain an intelligent integrated model of current efficiency. The simulation results show that the proposed model has high prediction accuracy and can be used in actual production forecast of current efficiency of aluminum electrolysis.