Using a Realization Technique for System Reconfigurability Evaluation: Simulation Application on a DC Motor

Tolerance to faults of control systems has been defined through the control reconfigurability concept, property defining the capacity of the control system to allow restoration of performance in the presence of faults. For its evaluation it is required to calculate the controllability gramian. This calculation is usually carried on before control system be implemented, i.e., off-line. Instead, the on-line evaluation of the control reconfigurability from input/output data when the system operates under given conditions (occurrence of faults), could be useful for the fault tolerance redesign, analysis and control system quality assessment in the actual operation. The contribution of the paper is to propose a data-driven way of calculation the controllability gramian on-line in order to evaluate an index based on the control reconfigurability. This index can help in supervision tasks for fault tolerance purposes. The paper aims to estimate, by using an identification algorithm, the reconfigurability using the controllability gramian of the system under partial loss of effectiveness type faults. An example is presented to illustrate the method.