This work refines the soft finger grasping fabric grabbing model and studies pure cotton woven fabric as an example to improve clothing item automatic grasping.
Fifteen types of pure cotton woven fabrics were selected for this study. And orthogonal experiments were used to determine the ideal soft finger grabbing parameter matching method (pressure and downward distance). The grabbing effect was then examined in relation to soft finger and fabric properties. Finally, step-by-step regression analysis was used to create a soft finger-grabbing prediction model based on fabric properties.
When manipulating pure cotton woven fabric, the ease of grasping is influenced by the pressure and downward distance exerted by soft finger. The effect of these two parameters on the effectiveness of the grab must be considered. And it is more important to choose the appropriate value of soft finger pressure. Conversely, there is an inverse relationship between fabric characteristics and the number of gripping layers. The structure design of soft finger has application limitations and needs to be optimised for fabric characteristics. The findings of stepwise regression analysis indicate that fabric weight, soft finger pressure and downward distance, fabric static friction coefficient, weft bending and weft elastic modulus have statistically significant impacts on the number of gripping layers. The prediction model based on this achieves a prediction accuracy of 100%, facilitating users to improve the application efficiency of the solution and demonstrating its significant predictive and reference value.
To hasten the research process of autonomous layer-by-layer grabbing clothing components stack, the soft finger grasping model is upgraded and the research reference is provided for different materials.
