基于支持向量机的高炉炉况诊断方法

BF Status Diagnosis Approach Based on SVM

现实中高炉炉况的征兆样本集是有限的,常规的基于经验风险最小化原则的方法的应用效果并不理想。支持向量机方法是针对小样本集分类问题提出的,具有很好的泛化能力,因此采用最小二乘法支持向量机进行高炉炉况诊断。通过仿真试验证实此方法具有很好的诊断效果。

The BF status evidential samples are limited and the approach based on empirical risk minimization principle doesn＇t work well. SVM （support vector machine） approach is aimed at solving classification problem with a small sample of training and has better ability for generalization,LSSVM approach is proposed to diagnose BF status. Finally the effectiveness of the approch was evaluated by MATLAB simulation.