NONLINEAR FLEXIBLE LINK ROBOT JOINT-FAULT ESTIMATION USING TS FUZZY OBSERVERS

Zahra Shams and Saeed Seyedtabaii

Keywords

Process fault, TS fuzzy observer, robust observer, robot joint-fault, fault estimation

Abstract

In this article, the observer-based fault diagnosis algorithms for the estimation of joint damage in an under-damped nonlinear flexible link robot are investigated. To have a convex objective function and use linear matrix inequality algorithms, the noisy nonlinear system is approximated by Takagi–Sugeno fuzzy linear model. The fault is defined by both the multiple proportional–integral (MPI) and algebraic fault models (AFMs). The results indicate that both methods estimate the measured states similarly while MPI renders more smooth process fault and unmeasured state estimates at faster tracking speed, but negatively, at higher computational costs. As a result, MPI is recognized to be a more attractive and robust, at least, in estimating the state/joint-fault of an under-damped robot system. The conclusion is validated through extensive simulations.

Important Links:

Go Back