Feedback Linearization Technique for DC Series Motor with LMI Based Observer and DREM

This paper presents a Feedback linearization based control strategy applied for DC series motor with Linear Matrix Inequality (LMI) observer. For using a closed loop control scheme for DC motor, it is necessary to identify parameters defining the electrical as well as mechanical dynamics. The system parameters are determined in open loop case using Dynamic Regression Extension and Mixing (DREM). The DREM procedure has a substantial advantage that the convergence of the unknown model parameters is ensured with improved transient response and finite time convergence property and it basically works on the non-square integrability condition on determinant of extended regressor matrix. These estimated parameters are updated in transformation as well as in observer model. The LMI is solved for optimal gain of the observer to maintain the closed-loop system stability. Simulation results are carried out for a typical motor operation scenario and demonstrate the efficacy of parameter estimation and state estimation on DC series motor.