An Adaptive Trajectory Planning Method for Acupoint Deviation Based on Improved Dynamic Movement Primitives

An improved trajectory-planning method for pestle-needle robots was developed for adaptive acupoint tracking and compliant surfaces fitting.A Gaussian-window-based attraction-time coupling mechanism was introduced to guide trajectories toward dynamic acupoints in both spatial and temporal dimensions.In two-dimensional simulations,the method reduces trajectory distortion by 16.35% and response time by 2s,alleviating deformation and rhythm disruption caused by acupoint shifts.An attraction-strength modulation strategy was designed to improve acupoint proximity while suppressing shape distortion,producing an additional 7% reduction in shape loss and improving path consistency.A curvature-inspired spring-damper model was applied along the vertical direction to improve surface fitting to curved back surfaces,reducing mean squared error by 21.5%,mean absolute error by 9.8%,and maximum error by 2.3%.Compared with RRT-APF,A^*-APF,Informed-RRT^*-APF,and D-ProMP,the method achieves higher trajectory accuracy,better stability,and smaller distance to acupoints,with an average minimum distance of 0.009m and a replanning time of 5.379s.CoppeliaSim simulations show that the method has good robustness,adaptability,and rhythm consistency.