Matlab Functions for the Grasp Force Prediction of Soft Robotic Fingers using Neural Networks

Matlab functions of three neural networks. Each of these networks predicts the grasp forces produced by one of three designs of soft robotic fingers. The networks were trained using FEA data and they produce results consistent with experimental data.

See the paper: "Bistable Stopper Design and Force Prediction for Precision and Power Grasps of Soft Robotic Fingers for Industrial Manipulation" by Xiaowei Shan and Lionel Birglen, accepted to the ASME Journal of Mechanical Design, 2023 for more details.

Soft fingers:

Neural network topology:

Experimental validations:

DISCLAIMER:

These files are provided "as is", and at your own risk. Under no circumstances shall any author be liable for direct, indirect, special, incidental, or consequential damages resulting from the use, misuse, or inability to use these files, even if the authors have been advised of the possibility of such damages.