LPV approach to friction estimation as a fault diagnosis problem

Abstract

The control of systems that involve friction presents interesting challenges. Recent research has focused on detailed modelling of friction phenomena as a very complex and difficult modelling challenge. However, the friction effects acting in a dynamic system can be viewed as actuator faults with time-varying characteristics to be estimated and compensated within a Fault Detection and Diagnosis (FDD) scheme, so that the limitations arising from the use of a friction model are obviated. This work is motivated by the utilisation of robust Linear Parameter Varying (LPV) estimation approach providing effective and robust fault estimation. The approach is illustrated using a two-link manipulator system with Stribeck friction. Results show that the time-varying friction forces on each joint can be simultaneously and robustly estimated through the online measurement of the varying parameters.