An on-line measuring fusion model of lumber moisture content based on data fusion algorithm

Lumber moisture content is a key parameter for regulating and controlling wood drying process. Its precision directly affects the drying quality, cost and drying time. In this paper a fusion model capable of on-line measuring lumber moisture content is presented. Models for predicting lumber moisture content are established using both back-propagation neural networks (BPNN) and dynamical recurrent neural networks (DRNN). Furthermore, the two models are integrated by arithmetic average and recursive estimation algorithm. The simulation result, which is worked out by experimental data , shows that fusion model have a higher predictive precision than any one of BP neural network´s and DRNN´s, therefore, this method is proved to be feasible.