A robust new nonlinear fault tolerant control for robot via an observer

To study the properties of the nonlinear robot model and introduce a new adaptive fault tolerant strategy, this paper has been introduced. The adaptive fault tolerant control scheme is designed via two (PID) Proportional -Integral - derivative controllers and a nonlinear observer. The nonlinear model of two inverted robot joints arms on a cart and has been studied and simulated with presence of an additive fault and a disturbance (white noise) in the system. The nonlinear adaptive observer is designed to eliminate the effect of the Coriolis and centrifugal torque and detect the fault and noise. The two PID controllers compensate the errors to increase the integrity of the strategy. The proposed strategy reconfigures the control law to overcome the errors and faults. An illustrated example is simulated and the encouraged results have been obtained.

I. Introduction

The fault tolerant control systems have been of interest to a lot of researchers over the past decades where the observers of the fault detection and diagnosis are main part of the fault tolerant control. For example, an extension of Luenberger observer, proportional and integral (PI) observers had attracted much attention during the past three decades. By assuming unknown disturbances to be constant, PI observers were studied in [1], , [3]. A robust (proportional, multiple-integral and derivative) PMID observer is designed to simultaneously estimate system states, fault signals, the derivatives of the faults, and attenuate disturbances successfully in [4]. A lot of FTC strategies rely on the observers for the fault detection and diagnosis. For a more detailed survey read [5], , , , , , [11].