基于聚类算法的仿生压曲锤头设计.

刘 权; 徐雪萌; 杨 磊; 唐静静; 徐永森; 李颍鹏

For the practical problems such as high labor intensity and low efficiency of manual treading, and the inability of traditional mechanical quilling technology to achieve the effect of manual treading, a bionic quilling hammer head was designed based on the clustering algorithm. Through the collection of plantar biomechanical data and analysis of plantar force in each region of the artificial treading, the force characteristics of the plantar in the treading movement were studied in depth. Based on the K-means clustering algorithm, combined with MATLAB software and Lagrange interpolation method for correction of outliers, the collected plantar biomechanical data were analyzed by clustering. The plantar region was divided into 4 categories, and the simulation results show that the SSE is 15 189.35, the CH score is 1 343.6, and the contour coefficient is 0.827 9, and the pressure characteristics of the same category after division are similar. The same category of plantar region after division was used as the pressure unit module to design the bionic koji press hammerhead in combination with the design. The bionic hammerhead design improved the area ratio of pulping effect by 5.94% and 3.19%, the internal uniformity factor by 3.9% and 1.07% compared to the flat and curved hammerheads. The study provides a reference for the development of pressing hammer heads.

K-means clustering; BIONICS; DATA analysis; INTERPOLATION; HAMMERS

Packaging & Food Machinery, 2024, Vol 42, Issue 5, p96

1005-1295

Academic Journal

10.3969/j.issn.1005-1295.2024.05.011