【作者】 师丽强； 张斌； 阳春华；

【Author】 SHI Li-qiang,ZHANG Bin,YANG Chun-hua(School of Information Science & Engineering,Central South University,Changsha 410083,China)

【机构】 中南大学信息科学与工程学院；

【摘要】 在净化除钴过程中,复杂的现场生产使工艺参数具有高维性、强耦合性、非线性等特点,对除钴效果分析以及过程建模产生较大影响。针对该问题,本文通过对现场工艺参数的特性分析,提出了一种基于核主元分析(KPCA)的数据预处理方法,消除噪音并有效提取工业数据信息。采用工业生产数据对该方法进行对比仿真实验,结果表明,基于KPCA的数据预处理方法改善了模型的精度,对净化除钴优化控制具有重要意义。更多还原

【Abstract】 Due to the complexity of the on-site production process,the raw data in Cobalt Removal of Zinc Purification process has characteristics of high dimension,strong coupling and non-linear,which affects the analysis of the cementation efficiency and modeling accuracy.Aimed to solve the above problems,a data pre-process method based on Kernel Principle Component Analysis(KPCA) is proposed after analyzing the characteristics of the raw data in the paper.The method could extract the useful information,eliminate the noise.and the nonlinearity The simulation based on the industrial process data shows that the method can improve the model accuracy and is significant for the optimization control of the cobalt removal process in zinc purification.更多还原