论文部分内容阅读
为了提高模型的泛化能力和精度,提出了一种基于混合核函数的支持向量机(SVM)建模方法。所提出的混合核函数由径向基函数和多项式函数加权组合而成,克服了支持向量机模型中单个核函数的局限性。并利用量子粒子群算法(QPS0)对惩罚系数、核参数以及混合权重系数进行综合寻优,求取最优化参数组合,从而提高模型的精度。采用锌湿法冶炼净化过程现场数据对建模的方法进行了测试,结果表明,所提出的混合核函数支持向量机模型具有较好的泛化性能和预测精度,预测结果满足现场工艺生产的要求。
In order to improve the generalization ability and accuracy of the model, a support vector machine (SVM) modeling method based on hybrid kernel function is proposed. The proposed hybrid kernel function is weighted combination of radial basis function and polynomial function to overcome the limitation of single kernel function in support vector machine model. The quantum particle swarm algorithm (QPS0) is used to optimize the penalty coefficient, kernel parameter and mixing weight coefficient to get the optimal combination of parameters so as to improve the accuracy of the model. The method of modeling was tested by the field data of zinc hydrometallurgy purification process. The results show that the proposed hybrid kernel SVM model has good generalization performance and prediction accuracy, and the prediction results meet the requirements of field process production .