Optimization of production parameters of particle gluing on internal bonding strength of particleboards using machine learning technology

Abstract The particleboard (PB) production is an extremely complex process, many operating parameters affecting panel quality. It a big challenge to optimize the PB parameters. of particle gluing have important influence on internal bond (IB) strength PB. In this study, using grey relation analysis (GRA) and support vector regression (SVR) algorithm, prediction model was developed accurately predict IB through processing in line. GRA used analyze relational grade between PB, variables were screened. SVR algorithm train 724 groups sample data six IB. tested with 181 sets experimental data. verified by data, values mean absolute error (MAE), relative (MRE), root square (RMSE), Theil’s inequality coefficient (TIC) 0.008, 0.017, 0.013, 0.014, respectively. results showed that performance nonlinear based GRA–SVR superior, can be real-time