【作者】 何鹏； 王雅琳； 桂卫华； 阳春华；

【Author】 He Peng,Wang Yalin,Gui Weihua,Yang Chunhua (School of Information Science and Engineering,Central South University,Changsha 410083,China)

【机构】 中南大学信息科学与工程学院；

【摘要】 氧化铝返料各成分受工艺过程中人为、迟滞等不确定因素的影响,混沌特性显著,本文将支持向量机引入返料成分混沌时间序列的动态预测。考虑到各成分动态序列的长度和峰值突变性的特点,提出了一种改进的支持向量机模型,首先改进支持向量机模型的泛化函数,接着针对支持向量机算法存在的参数优化、训练和测试速度等问题,采用混沌优化算法优化SVM模型中的参数。改进的SVM模型用于氧化铝返料成分混沌时间序列预测,经生产数据验证具有较高的预测精度。更多还原

【Abstract】 Alumina returns compositions are affected by uncertainty factors such as human,lag and so on,their chaos characteristics are obvious.So support vector machine is introduced into returns compositions’ dynamic prediction.Considering the dynamic sequence’s length and characteristics of peak mutability,an improved support vector machine is proposed.First,the support vector machine(SVM) model generalization function is improved.Then chaos optimization method is applied to optimize the parameters of SVM model for solving the problems such as parameter optimization,training and testing speed and so on.The improved SVM model is used to predict chaotic time series of alumina returns compositions and its high prediction precision is verified.更多还原