Continuous route planning method based on density grading for thinning of walnut male flowers

Walnut male catkins commonly appear in clusters and are irregularly distributed across the tree, which makes it difficult to generate a continuous path during mechanical thinning. To reduce energy consumption and simplify route planning, a method termed continuous route planning of extensive and efficient mechanical thinning for walnut male flower based on density grading (CRPE′-DG) is proposed via a sub-regional route-planning strategy along the main branches. This method incorporates and improves the Density-Based Spatial Clustering of Applications with Noise (DBSCAN), B-Spline fitting, and Makima interpolation algorithm to generate end-effector motion routes for the robotic arm. The performance of CRPE′-DG is evaluated in terms of dispersion degree, thinning rate, and route length. The Interquartile Range (IQR) of the dispersion of male flowers at high-density levels is consistently lower than that at low-density levels. On average, the dispersion degree is reduced by 57.5% and 31.9% compared with the 5th-order and 7th-order polynomial fitting algorithms, respectively. The average thinning rate using the CRPE′-DG method has been increased to 83.6%, which is 11.3% and 6.3% higher than the 5th-order and 7th-order algorithms. Meanwhile, the average route length is shortened by 15.3%, 18.0%, and 14.1% compared with the 5th-order, 7th-order, and traditional B-spline algorithms, respectively, with a maximum reduction of 28.0% in sample 1. These results demonstrate that CRPE′-DG can effectively improve thinning efficiency, generate more continuous routes, and reduce energy consumption, thereby supporting sustainable agricultural development.