Predicting the Mechanical Properties of Heat-Treated Woods Using Optimization-Algorithm-Based BPNN

This paper aims to enhance the accuracy of predicting mechanical behavior wood subjected thermal modification using an improved dung beetle optimization (IDBO) model. The IDBO algorithm improves original DBO via three main steps: (1) piece-wise linear chaotic mapping (PWLCM) generate initial species and increase its heterogeneity; (2) adopting adaptive nonlinear decreasing producer ratio model control number producers boost algorithm’s convergence rate; (3) applying a dimensional learning-enhanced foraging (DLF) search strategy that optimizes ability explore exploit space. is evaluated on 14 benchmark functions outperforms other algorithms. then applied optimize back-propagation (BP) neural network for five property parameters heat-treated larch-sawn timber. results indicate IDBO-BP significantly reduces error compared with BP, tent-sparrow (TSSA)-BP, grey wolf optimizer (GWO)-BP, grouping (IGWO)-BP DBO-BP models, demonstrating superiority in physical characteristics lumber after heat treatment.