Ncut聚类与增量支持向量机的SMB智能建模

SMB intelligent modeling based on Ncut clustering and incremental support vector machine

针对当前SMB(simulated moving bed)难以实时在线测得输出组分纯度的现状,结合Ncut(normalized cut)聚类及增量学习支持向量机的方法建立达到周期性稳定状态时系统的智能模型。采用Ncut方法对离线采集的数据样本进行聚类,得到样本的聚类结果;将聚类后的样本数据按反复记忆增强机制输入向量机进行增强-增量学习训练;将原始测试样本输入到训练好的模型中进行检验。检验结果表明,采用该模型可以获得更好的模型适应度和检验精度,仿真结果验证了该方法的有效性。

The current SMB（simulated moving bed）is difficult to measure the purity of output component online.To solve the problem,the Ncut clustering was combined with the incremental learning of support vector machine when the system is stable to build black box model.The method of Ncut was used to collect off-line data samples to cluster and the clustering results of training samples were got.The results were used to enhance the incremental training according to repeated memory enhancement mechanism of input vector machine.The original test sample was inputted into the trained model to test.The results show that combining the Ncut clustering with the incremental support vector machine model can get better model fit and test precision,the results of simulation verify the effectiveness of the method.