Related Experiment Video
Updated: Jun 24, 2025
09:41
Estimation of Contact Regions Between Hands and Objects During Human Multi-Digit Grasping
Published on: April 21, 2023
1.6K
Real-time haptic characterisation of Hunt-Crossley model based on radial basis function neural network for contact
Jiankun Li1, Xinhe Zhu1, Yongmin Zhong1
1School of Engineering, RMIT University, Melbourne, VIC, 3083, Australia.
Summary
This study introduces a new method combining neural networks and recursive least square (RLS) estimation for accurate dynamic soft tissue characterization in robotic surgery. The approach improves the estimation of Hunt-Crossley (HC) model parameters, enhancing surgical precision.
Area of Science:
- Robotics
- Biomedical Engineering
- Computational Mechanics
Background:
- Accurate dynamic soft tissue characterization is crucial for robotic minimally invasive surgery.
- Existing models often face challenges with nonlinearities and estimation errors.
Purpose of the Study:
- To develop a novel method for dynamic soft tissue characterization using neural networks and recursive least square (RLS) estimation.
- To improve the accuracy of the nonlinear Hunt-Crossley (HC) model parameters estimation.
Main Methods:
- A radial basis function neural network (RBFNN) was developed to compensate for errors from natural logarithmic factorization (NLF) in the HC model.
- RBFNN weights were estimated using the maximum likelihood principle.
- An RBFNN-based RLS algorithm was created to compensate for linearization errors.
Main Results:
- The proposed method effectively models the natural logarithmic linearization error of the HC model.
- Improved accuracy in RLS estimation of HC model parameters was demonstrated through simulations and experiments.
Conclusions:
- The RBFNN-based RLS method offers enhanced accuracy for dynamic soft tissue characterization.
- This advancement has significant implications for improving control and safety in robotic surgery.
