论文部分内容阅读
针对传统支持向量机和单一模型建模的缺点,利用某炼油厂溶剂油分离过程中二侧线流量作为建模对象,对最小二乘支持向量机集成学习方法进行了研究。首先利用自适应系数加权模糊(AWFCM)聚类算法对训练样本进行聚类;然后对每一类数据使用最小二乘支持向量机建立子模型,并使用PLS合成函数得到最小二乘支持向量机集成模型;最后通过仿真实验来验证最小二乘支持向量机集成模型预测的精确性。结果表明,该算法在预测精度上有了较大的提高,对过程控制系统中分离效果的预测具有重要指导意义。
Aiming at the shortcomings of traditional support vector machine and single model modeling, the method of two-sided linear flow in the process of solvent oil separation in a refinery was used as modeling object, and the integrated learning method of least square support vector machine was studied. Firstly, the training samples are clustered by the adaptive coefficient weighted fuzzy (AWFCM) clustering algorithm. Then, for each type of data, the least squares support vector machine is used to build the submodels and the PLS synthesis function is used to obtain the least square support vector machine Finally, simulation experiments are carried out to verify the accuracy of the integrated model prediction of least square support vector machines. The results show that the algorithm has greatly improved the prediction accuracy and has important guiding significance for the prediction of the separation effect in the process control system.