Study On Larch Wood Drying Models By Artificial Neural Network

The wood is the most important raw material resources in daily living and production, the demand for wood is increasing every day. Efficiently utilizing wood and improve quality utilization of timberwork are already became one of the most main problems which is earned more and more attention by people. Wood drying is the key technology of reduce the wood loss or increase the woodwork quality, and the practical significance of reduce waste national resource also based on rational wood drying. The wood moisture is an important index in the wood drying process; it will directly affect stability of wood drying. Only by reduced the wood moisture to a certain degree that we can meet the need of quality in the process of wood drying, it is also the final task. The research of wood moisture model is reflected in drying schedule. From all the wood drying control models considered, they all from one point of view reflected the change trend of moisture content. The research on prediction wood moisture by drying internal environment factors has practical significance on reduce the drying time.This paper determines wood moisture of the larch of Russia which is belongs to the Northeast Forestry University’s laboratory by using BP neural network three-layer mapping model. Take the dry bulb temperature, equilibrium moisture content measured by dryer internal, wood moisture during preheating period and drying stage as reference obtain the input matrix; then take the wood moisture during cooling phase and damp-heat stage obtain the output matrix. By this method, determined three-layer mapping and got the result that the optimal interlayer was ten through multiple tests, then take ratio5:10:1to predict moisture of cooling and damp-heat stage. In order to control the temperature during drying process by the prediction of moisture content through adjust the dry bulb temperature and other parameters so as to made the wood moisture is controllable in semiautomatic controls. Then got the parameters of wet and dry bulb temperature, wood temperature, wood equilibrium moisture content by using multiple regression analysis and got the regression coefficient was0.88of optimal regression equation.