Identification of friction at low velocities using wavelet basis function network
Hongliu Du, Satish Sasidharan Nair · 1998
Modeling friction at low velocities can be viewed as a problem of characterizing the local behavior of a complex nonlinear function. Wavelet theory provides a very powerful framework for localized mapping and learning. Based on space-frequency localization, a model-free methodology to identify and compensate for friction at low velocities using a wavelet basis function network (WBFN) based design is proposed and validated using a hardware example case. The methodology is systematic and uses minimal knowledge of the dynamics which is particularly attractive for a large class of dynamic systems with friction.