FDI based extended kalman filter for multi robot system

Abstract

In this paper the problem of detection and isolation (FDI) in wheeled mobile robot is treated. first of all we consider a master-slave robot system in which the master robot has to follow a desired trajectory given by a well-determined control law, then the slave is controlled by the master. However in real environment there is no ideality because faults can appear in the master or slave or even the both and affect the system and may lead to unacceptable results so to ensure a well functioning an extended kalman filter is tuned to diagnosis eventual faults and to generate residuals in order to detect and isolate them. The performance of an estimator based on an EKF is simulated based MATLAB Simulink simulations.