Rapid Detection of Formaldehyde Emission from Wood-Based Panels Based on the IPOA-XGBoost

Aiming at the complexity of the process of analyzing formaldehyde composition by physicochemical methods in the process of formaldehyde emission detection by the full-scale chamber method, and the problem of bias in the accuracy of sensor method, the IPOA-XGBoost formaldehyde emission modified deviation model for a class of panels is proposed for the fast and accurate detection of formaldehyde emission from wood-based panels. The model utilizes data obtained from the electrochemical sensor as input. The output is the formaldehyde emission measured by the full-scale chamber method. The model is fitted using the extreme gradient boosting (XGBoost). To address the weak optimization ability in the pelican optimization algorithm (POA), random opposition-based learning is employed to establish the initial population. Additionally, nonlinear weighting factors and a sparrow alert mechanism perturbation strategy are introduced to enhance the ability of global optimization and escape local optimal solution. The improved POA (IPOA) is used to optimize the key parameters of XGBoost, and then the modified deviation model of formaldehyde emission of the IPOA-XGBoost wood-based panel is constructed, and compared with four machine learning models. Experimental results indicate that the proposed model achieves a coefficient of determination of 0.9701 and a root mean square error of 2.7648e-03. Furthermore, it exhibits lower prediction errors compared to other models tested herein. This provides an effective and reliable solution for rapid detection of formaldehyde emission in wood-based panels.