A simulation and experimental study on identification of a DC motor

The current applications in electromechanical energy conversion demand highly accurate speed and position control. For this purpose, a better understanding of the motion characteristics and dynamic behavior of electromechanical systems including nonlinear effects is needed. In this thesis, suitable model of Permanent Magnet DC (PMDC) motor is developed for identification purposes. Model is parameterized and identified via simulation and using real experimental data. Linear and nonlinear models for the system are built for identification, and the effective nonlinearities in the system, which are Coulomb friction and dead zone, are integrated into the nonlinear model.A Weiner-Hammerstein nonlinear system description was used for identification of the model. MATLAB is selected as the investigating tool, and a simulation model is used to observe the prediction error between the simulated and estimated outputs. Identification of the linear and nonlinear system models using experimental data is performed using the least squares (LSM) and recursive least squares (RLS) methods. Performance of the model and identification method with the real time experiments are presented numerically and graphically, revealing the advantages of the proposed nonlinear identification approach.Key Words: Nonlinear system identification, Wiener-Hammerstein model, Nonlinear PMDC motor.