Advanced hybrid metaheuristic framework for spiral rubber tapping trajectory optimization on curved tree trunks

Automated rubber tapping requires accurate reconstruction of smooth and kinematically feasible trajectories from discrete tapping points on complex curved trunk surfaces. However, achieving a balance between geometric smoothness and fitting accuracy remains challenging. This study proposes an optimization-driven trajectory modeling framework that integrates an Adaptive RIME algorithm with Bézier curve parameterization (ARIME-Bézier). While Bézier curves inherently ensure geometric continuity, their performance strongly depends on polynomial degree and control point configuration. To overcome this limitation, the proposed Adaptive RIME incorporates an adaptive hard-rime puncturing mechanism and a time-varying step-size strategy to mitigate premature convergence. The framework jointly optimizes the polynomial degree and control point distribution, enabling a balanced trade-off between statistical fitting precision and kinematic smoothness required for stable robotic tapping. Experimental validation is conducted on a standard spiral tapping path (32° slope angle, 80 mm trunk radius, 180° wrap angle). Comparative results demonstrate that the proposed method consistently outperforms baseline approaches, achieving R² > 0.90 across all cases. The model attains a root mean square error of 5.9137, with average radial and slope angle deviations of 0.8% and 1.74%, respectively. These findings confirm the robustness and practical applicability of the proposed framework for intelligent robotic rubber tapping systems.