Study of PMSM drive fed by PWM Inverter using software.

There are various types of permanent magnet synchronous machines, PMSM. Here we present two types; the brushless DC machine and the interior permanent magnet synchronous machine.

The brushless DC machine is the most simple type and easy to control. It is mostly equiped with a position sensor, form which the gate contorl signals are derived. Since the brushless Dc machine operates with constant on-time square wave voltages, only discrete postion signals are required and mostly three Hall devices are enough. In the simulation below the three hall sensor signals are converted into the gating signals using the block (Commutation sequency 120 degrees) that directly controls the mosfets in the inverter. Since this is a constant on-time control, there is no speed control nor current control. Only hte DC link votlage can be controlled to control the final drive speed.

The simulation shows the current through the DC link and the machine currents. Next is the voltage between phase a and phase b, showing a trapezoidal waveform. The nothes in this waveform are due to the switching of the mosfets in the inverter. The torque is displayed and its shape follows the shape of the DC link current and th eamplitude of the pahse currents. The Hall signals h1 h2 and h3 are the outputs from the Hall position sensors, from which the gating signals g1 g2 g3 l1 l2 l3 are derived.

The brushless DC machine has trapezoidal back emf and on the right side of the model, the electrical position of the rotor is exported. This signal can be used for simulation, but is in reality not available. By enabling the animation, one can see the switching in the inverter as response of the Hall signals.

Click to close the image

The interior permanent magnet synchronous machine with inverter and field oriented control is used in speed control and traction. It is found in many electric vehicles. Compared to the brushless Dc machine, the back emf is sinusoidal and the winding inductance can be defined as Ld and Lq. Saturation may be added via the bottom connections of the machine model. Harmonics can be inluded in the back emf waveshape to model various magnet geometries. At the top of the machine model there are connections for a thermal model.

The inverter with modulator are modeled as one averaged model block. The overall advantage of this model is its simulation speed, allowing designing hte current controller within reasonable time. The top connections of the inverter model are the connections to a thermal model of the heatsink.

The field oriented control is build using Clarke, Park, PI controllers, inverse-Park and spave vector modulation blocks. The PI controllers are discrete components, so also the sampling frequency should be defined here.

The measured currents are transformed into ia and ib using the Clarke transformation. Using the Park transformation and position of the rotor, they are transformed into rotor refence coordinates id aand iq. Simply said, alpha and beta are somehow sinusoidal, while dq are more or less dc values.

The information from ia, ib, id and iq is used in the animations. The left animation shows the rotating stator field and rotating rotor, as if we are sitting on the stator. The right animations shows us the stationairy stator currents and fixed rotor position as if we aer sitting on the rotor and rotating with the angular rotor speed.

In this simulation only a current control is modeled. Torque is produced by the current iq, the field can be weakened by negative id.

Click to close the image

Both surface mount and interior permanent magnet motors can be modeled. Also axial permanent magnet machines can be modeled using the BLACM machine model. Here the correct vlaues of Ld and Lq as function of id and iq as well as the harmonic content of the back emf waveform can be defined, allowing the modeling of almost any type of permanent magnet synchronous machine.


© 2017 CASPOC, All rights reserved. Home   |   Terms and Conditions   |   Legal   |   Export Compliance   |   Privacy Policy   |   Site Map