2014
MAS 2014 Proceeding
Supervised growth network SGWRN model can achieve a balanced control of two-wheeled robot
Authors: Fu-jin Li, Mei-jie Huo, Hong-ge Ren, Dong-mei Fei
Abstract
Based on the unsupervised growth structure of the network GWRN model, GWRN model and radial basis function network architecture combined. Proposed supervised structure of the network can be grown SGWRN models. Model used is the growth of self- organized growth algorithm when needed, by inserting new neurons, changing the connection among competition layers and adjusting the connection weights between layers, achieve the required accuracy of the model output, simulation experiments show that, SGWRN model realizes the two balancing robot control, there are certain anti-interference ability and practicality.
I3M Scientific Sponsors
I3M Industrial Sponsors
I3M Media Sponsors
