System identification using experimental data acquired for a few values of inertia and approximating the influence of inertia using interpolation

The system identification method proposed in this paper has been used to develop a simulation model for a system composed by a DC motor with a variable inertia and a current regulator (regarded as a black box). This method performs multiple SISO (single input - single output) identifications of the system, each time with a different constant value of inertia. Using all SISO identifications, a single model of the system is developed using state space averaging to perform the interpolation, in order to model the behavior of the system for intermediate values of inertia.