LAUSR

Lumber is an indispensable raw material for people ’ s daily life. Drying process is a crucial stage in lumber manufacture. As to ensure a suitable and usable end product of lumber, most of its MC must be removed by drying. Improving the quality of lumber drying requires efficient control scheme. This article presents a design of radical basis function (RBF) neural network–based sliding mode controller for lumber drying system. RBF neural network is introduced to optimize the conventional slidingmode controller. The proposed strategy has been theoretically and experimentally investigated to demonstrate the applicability for lumber drying process. Comparative study of conventional sliding mode control (SMC) scheme and proposed control scheme is also presented. It was found that the control performance of RBF-based sliding mode controller was superior to the conventional slidingmode controller in computer simulation. Furthermore, in the field conditions, time and energy consumption reduction were noticed with RBF-based sliding mode controller compared with conventional SMC strategy using the same drying schedule, although the drying quality using the two control methods were similar.