Near-time-optimal position control of an actuator with PMSM

Abstract

The paper presents design and implementation of 'near-time-optimal' position control of electrical drive employing permanent magnet synchronous motor. The position control algorithm, which is applied to the motor, is derived from a simple double integrator model. Then principles of vector control, forced dynamics control and time-optimal control are combined to form suitable control strategy, which was developed as two modifications. The first developed control strategy is model based. Near-time-optimal double integrator model, which works in real time computes 'near-time-optimal' response to the demanded position. This computed position then provides the reference input for the drive position control loop. To compensate some delay between generated position in the model and real drive position the precompensator was implemented. The second modification of the control strategy is based on direct computation of the stator current torque component from the known required torque of the machine. Load torque estimate, which is necessary for both control strategies yields linear third order observer. Comparison of these two control strategies based on experimental results is given for an actuator with PMSM. Presented experimental results for model based and direct torque component computation of NTO control direct show possibility to achieve ''near-time-optimal' behavior of the drive employing PMSM for both control strategies