A novel multi objective constraints based industrial gripper design with optimized stiffness for object grasping

Abstract

Soft gripper design is a rising area of research due of its great possibilities in automation. One difficult problem in robot design is the ability to grasp a broader variety of items with variable stiffness, forms, and sizes in a single gripper. An ideal soft robotic gripper design with variable stiffness was designed in this research as a grasping model. Its distinctiveness is found in the methods utilized for modelling actuators and in the shifting stiffness characteristics of silicon soft gripper. When modelling the actuator in this case, multi-objective functions like gripping displacement and force transmission ratio are taken into account, and the actuator functions are controlled by the MDF (multiple degrees of freedom). The precise stiffness needed to grasp the item is then chosen using an adaptive optimization method. This enhanced weight-based horse herd (IHH) optimization method carries out the stiffness adjustment based on actuation pressure. Additionally, the suggested soft robotic gripper with variable stiffness employs the adaptive level set (ALS) technique to build the gripping force model. Additionally, several validations are offered in relation to the outcomes for item gripping by the suggested soft gripper. This shown that the results of the created soft gripper excelled those of other methods. The developed ABBIRB 1410 robot gripper type may enhance the work cycle in industrial applications and performs object grabbing with dependability and speed. The experimental validations show that the developed gripper model provides an enhanced object grasp with a range of curvatures, delivering a maximum pulling force of 121.07 kPa at 50 kpa, 119.15 kPa for patterned pulling, and 45.05 kPa for non-patterned pulling. The designed gripper type has a curve with a minimum size of 1.1 mm and a maximum size of 218 mm. Additionally, the soft gripper for industrial applications is examined with variously sized and weighted items. The suggested gripper model achieved an RMSE performance of 2.9 and a Pearson correlation of 0.993.