关键词: 支持向量机, 软测量, 质量控制, 建模

Abstract: On the basis of studying Support Vector Machine (SVM) theory, a soft-sensor controlling method based on Support Vector Machine wass presented. In order to solve the problem of getting the important parameter that is hard to be measured online and has long time-delay, a soft-sensor controlling method based on support vector machine was presented. In the control process, modeling techniques have been studied intensively, and then RBF kernel function was chosen to establish an exact support vector machine model. On the background of quality control in a company, the online estimate of output value was realized. Under the circumstance of changing and choosing different parameters and through a lot of research and simulation, a relatively better generalization result model was established.

Key words: Support Vector Machine (SVM), soft-sensor, quality control, modeling