Research on GA-BP neural network model of surface roughness in air drum sanding process for poplar

Roughness is an important property of wood surface and has a significant influence on the interface bonding strength and surface coating quality. At present, there is little research on theoretical models for poplar fine sanding. In this study, poplar wood was fine-sanded with an air drum. An orthogonal experiment was carried out to study the effects of abrasive grain size, feed rate, belt speed, air drum deformation and air drum pressure on the surface roughness of poplar wood. The simulation models of the longitudinal roughness and the lateral roughness of the sanding surface were established based on BP neural network optimized by genetic algorithm (GA-BP neural network), and verified by the experimental data. The results show that the influence of sanding parameters on longitudinal roughness and lateral roughness is similar. The order of influence is abrasive grain size > belt speed > feed speed > air drum deformation and air drum pressure. The longitudinal roughness and lateral roughness of the surface of the poplar can be well predicted by GA-BP neural network. The average relative error of the predicted longitudinal roughness and lateral roughness are 2.67% and 2.65%, respectively.