MOISTURE CONTENT PREDICTION OF WOOD DRYING PROCESS USING SVM-BASED MODEL

摘要

In this paper, a moisture content predictive method is proposed for wood drying process by using nonlinear model based on support vector machine. In detail, the properties of the moisture content are analyzed during wood drying, which shows that the moisture content is mainly determined by drying temperature and equilibrium moisture content and that water in wood consists of two forms: free water and bound water. For ensuring high predictive accuracy, a moisture content predictive model related to drying temperature and equilibrium moisture content is built by using support vector machine technique. Further, practical parameters selection of SVM modelling is investigated. Also, to filter large noise of training data, infinite impulse response filter technique is considered. By using the built model, moisture content of wood drying process is predicted, including free moisture content and bound moisture content. Finally, simulation results are given to show the effectiveness of the proposed method.