Cold rolling process data analysis based on SVM

With the characteristic of simplification and low-precision, conventional rolling production can´t satisfy the requirement of high-precision rolling, although self-adaptive technology is used in actual processing. To improve the rolling precision and reduce the cost, rolling model analysis framework based on data is presented by using the statistical learning theory (SLT) and support vector machine (SVM). SLT is the basic theory and mathematical framework that researches small-sample data. As one of the advanced intelligent algorithms based on SLT, SVM is used to regress the rolling force parameter in the paper. The conclusions indicate that SLT and SVM are applicable to research industrial process like rolling production.