Stochastic modelling and identification of lubricated polymer friction dynamics

We apply a maximum-likelihood stochastic system identification technique to the problem of identifying parameters of a model for friction in a hydraulic actuator. This is the first application of stochastic modelling and identification techniques to the problem of modelling friction, which is a very complex physical process. The deterministic part of the model characterizes energy dissipation mechanisms of friction and the associated transient responses. The stochastic part of the (ARMAX) model characterizes unmodelled dynamics due to process disturbances and measurement noise. The model and identification algorithm are validated by comparison with experimental data.