摘要
以浮选过程关键控制量(浮选加药量)为研究对象,提出一种基于核主元分析(KPCA)和最小二乘支持向量机(LSSVM)的浮选加药控制模型。通过对浮选过程的工艺的分析,确定了模型的输入输出变量。采用KPCA算法对样本数据进行降维处理,简化模型结构,并用LSSVM建立浮选加药控制模型。采用基于高斯变异和柯西变异改进的混合蛙跳算法(ILSFA)对LSSVM模型的参数进行优化,获得高精度的控制模型。仿真结果表明,提出的模型能够显著提高出浮选加药的控制精度,满足浮选加药过程的需求。
For controlling the key control variable （the flotation dosage） of the flotation process, a control model based on kemel principal component analysis （KPCA） and least squares support vector machine （LSSVM） was proposed. This paper determined the input and output variables through the analysis of the flotation process, simplified the model structure by dealing with the sample data with KPCA and then established the flotation dosing control model with LSSVM. The shuffled frog leaping algorithm was improved with Gauss mutation and Cauchy mutation method and the optimal parameters of LSSVM model were searched by the ISFLA. The simulation results show that the proposed model can significantly increase the control precision of the flotation dosage and can meet the needs of the flotation dosing process.
出处
《控制工程》
CSCD
北大核心
2017年第2期326-330,共5页
Control Engineering of China
基金
国家自然科学基金(61473054)
