An EKF-based algorithm for automatic tuning and control of electrical drives

The paper presents a DSP-based procedure for the automatic tuning of speed and position controllers of electrical drives. The procedure is based on an identification algorithm (namely an extended Kalman filter) for the determination of the parameters of the system model, assumed as a first-order lag defined by its mechanical constant. The algorithm, which is executed at the start-up of the drive, transfers its results to another procedure which exploits the identified parameters for the optimal tuning of the drive motion controllers. The tuning is carried out off-line in discrete time. Theoretical and experimental results of the chosen tuning procedure are reported. For the design and testing of both the identification and regulation algorithm a simple PC-based tool has been developed, which provides the hardware support to command different actuators, together with the required software facilities.<
>