A non-parametric friction model for accurate positioning control using v-support vector regression (v-SVR)

Motivated by the need for simple and efficient friction model required for precision positioning control, a non-parametric-based technique using v -support vector regression (v-SVR) is proposed and developed in this work for the modeling of friction in a DC motor-driven motion control system. The effectiveness of the developed friction model to compensate the frictional effects is evaluated experimentally on the system for both point-to-point (PTP) and tracking positioning control. The performance of the model was benchmarked with a parametric model. The results show the proposed model as a viable and efficient alternative to mathematical-based techniques in representing and compensating friction effects.