Comparative Analysis of Three Methods for Stochastic Lumber Drying Simulation

Abstract: This article describes a novel stochastic model designed to simulate systems that cannot be analyzed as a unit, but as a collection of a large number of similar components. In order to state advantages and disadvantages, the proposed method is compared with two other published models. The first is a symbolic mathematical relationship designed to predict average moisture content and standard deviation after conventional drying of lumber. Since this model is exact, it was used as reference to evaluate the accuracy of the other approximate numerical methods. The second model is entirely random, and it emulates a real system behavior in which the parameters and conditions randomly change from one component to the other. The proposed method is based on numerical integration of the parameter's frequency distribution curves, which always produce the same and most probable result for the same parameters and conditions. The three methods were applied for simulation of conventional lumber drying, and the results were compared both qualitatively and numerically.