基于延展性的机器人面料缝制张力预测方法

针对机器人缝制过程中未知面料期望张力导致面料变形的问题,根据面料的延展性分类训练支持向量机(SVM)模型,通过线性支持向量机模型预测未知面料的延展性;其次,采用模糊逻辑控制系统确定面料特性与期望张力之间的非线性关系;最后,利用毛呢、绸缎、天鹅绒布、摇粒绒布4种面料对SVM模型进行测试,通过模糊逻辑得到面料期望张力查询表,根据期望张力对毛呢和竹节棉麻进行缝制实验。结果表明:在拉伸面料的过程中,线性支持向量机模型预测的延展性最终趋于面料实际的延展性,基于模糊逻辑根据面料延展性和种类可实现对任意面料期望张力的预测,预测张力可满足智能化缝制加工的需要。该研究为避免面料形变从而提高缝制质量提供了前提条件。

Prediction method for tension of fabric sewn by robot based on extensibility

In order to better understand the effect the sewing tension on fabric sewing flatness in the robotic sewing process, the support vector machine (SVM) model was trained according to the extensibility of the fabric, and the extensibility of a new fabric was predicted by the linear SVM model. A fuzzy logic control system was used to determine the nonlinear relationship between the cloth characteristics and the applied tension. The SVM model was tested with four fabrics, i.e., wool, silk, velvet, and flannel, among which silk was selected to output the expected tension of the fabric through fuzzy logic relations. The results show that in the process of fabric stretching, the extensibility predicted by the linear SVM model eventually tends to converge towards the actual extensibility of the fabric. Based on fuzzy logic, the expected tension of any fabric can be predicted according to the extensibility and fabric type. This research provides a prerequisite for avoiding fabric deformation and improving sewing quality.