基于LS-SVM的棉针织物染色上染率模型的研究

Study on the model of the uptake rate in dyeing of cotton fabric based on LS-SVM

针对棉针织物染色上染率建模中存在非线性建模困难及预测精度不高等问题,结合棉针织物的染色机理,基于最小二乘支持向量机(LS-SVM)建立上染率模型并进行预测.将提出的方法与BP神经网络法进行比较,试验结果表明前者具有更高的预测精度,验证了LS-SVM对织物染色上染率预测建模是一种可行而有效的方法.

The uptake rate model of cotton fabric based on least squares support vector machine（LS-SVM）is proposed in accordance with the dyeing features of cotton fabric,for it is difficult to develop a non-linear model for predicting the uptake rate and get a high-prediction accuracy.Comparing the proposed method with BP neural network based method,test results show that the proposed method is better in prediction accuracy and effective for modeling of the uptake rate.