Lumber Moisture Content Sptail Distribution Model Based Fusion Methods During The Drying Industry Process

During drying process, lumber moisture content is the guarantee of wood dimension stability and wood products quality key parameters. According to the statistics, the main reason of wood appear quality problem is because of moisture content detection is not accurate in wood drying stages. Research drying moisture content of wood drying kiln distribution characteristics is helpful to regulate drying parameter, adjust drying process and improve the finished product of drying products. At present, for timber moisture content data limit of the few points that can be measured in wood drying kiln, there are the problems that we can't describe the space moisture content distribution state in drying furnace, moisture transfer and change trend completely and accurately. Therefore, it have certain research significance that using discrete points data of wood drying kiln to set up a continuous and visual lumber moisture content space model and then analyzes the drying process of the moisture content of spatial distribution characteristics.In this paper, contra poses the problems of we can't describe the space moisture content distribution state in drying furnace completely and accurately, using fusion theory combining spatial analysis method to set up lumber moisture content space model.(1) Constructing lumber moisture content distribution functions according to experimental data. According to the function, respectively compared based on least squares fitting nonlinear method, Lagrange interpolation method, support vector machine forecasting algorithm and support vector machine (SVM) method based on particle swarm optimization algorithm. Simulation results show that support vector machine method based on particle swarm optimization algorithm used in wood space discrete point data to predict is of high precision.(2) Put temperature information and relative humidity information in the space into lumber moisture content input variables, and then use the output variable of estimates of wood moisture content input to mode of based on least square method of nonlinear curve fitting. Make the expansion of space and training with the data of discrete measurements of material in drying furnace. Set up support vector machine equation based on particle swarm optimization algorithm. Get spatial data on continuous lumber moisture content, to create the one-dimensional moisture content distribution model.(3) By creating a compensation coefficient r, on the predictable one-dimensional model to the expansion of training for space and applied single test materials on timber moisture content distribution function to the plane of wood drying kiln. Through experiment, the average error of this predictable method function is 1.44%. The method of two-dimensional interpolation was used to establish the consequent model of lumber moisture content in two-dimensional plane modeling. Finally, the kiln of wood space model mesh was constructed with method of 3D data visualization and Cartesian grids. In MATLAB7.0 environment, the kiln of wood space model in the drying process was established.Establish lumber moisture content model, the spatial distribution model using less wood moisture content sensors to intuitive showed moisture content space distribution and spatial moisture content trend. Improve drying process for material changing trend of space moisture content, reducing the prediction accuracy of wood drying process of losses in. At the same time, t studies using discrete points established content for continuous space model provides research methods, to strengthen and improve wood science and technology research level has certain simulative effect.