Radical basis function neural network for a novel tactile sensor design

In this paper, a decoupling algorithm based on the RBF (radical basis function) network is proposed. Compared to the Levenberg-Marquardt (L-M) algorithm and CMA-ES algorithm, numerical experiments show that the decoupling algorithm based on the radial basis function network is of higher performance on both accuracy and time complexity, which makes the real-time property of the tactile sensor possible. In the further research, the mapping between the force and the deformation of the tactile sensor surface can be revealed by the ANSYS simulation, and the RBF network plays a key role in the digital model of the tactile sensor.