E-learning Resources in Microelectronics Field Orientated Control of Induction Machines

Universität Paderborn

Document Information
 

1.  General

Author:

Staff of Universität Paderborn

1. General
 

Some years ago only separately excited DC machines were able to meet high dynamic demands.
Variable speed induction machines fed with normal frequency inverters allow only slow changes in speed because of the complex transient behaviour calling for high currents in the case of fast frequency changes.

Controlling a DC machine is much easier because of the fact that the main flux and the armature current distribution are fixed in space and can be controlled independently while with an AC machine these quantities are strongly interacting and move with respect to the stator as well as the rotor. They are, in a complex way determined by the amplitudes, frequency and phases of the stator currents. The three stator currents can be reduced to two independent control variables. With an induction machine, only the stator currents can be controlled and with an ordinary squirrel cage it is not possible to measure the rotor currents. Since the induction machine is a highly non-linear interacting multi-variable control plant, control engineers puzzled about this for a long time.

The solution uses the principle of field-orientated control, giving the induction machine the operating characteristics of a separately excited DC machine.

The following screen-shots show a PC-software for simulation, monitoring and parameter-set-up of an inverter for field-orientated control (david, www.d-tech.de).

The simplest mode of operation of the david inverter is V/f control, normally only used for activating a drive system for the first time.

With V/f control, the voltage of the motor is dependent on the frequency defined by the V/f characteristic curve.
The voltage is applied to the motor and the current depends on the motor state.

In field-orientated vector control mode, the currents are controlled, depending on the system state.
The following screen shot shows speed vector control with an encoder for position and speed measurement. Currents and voltages are measured by sensors and transformed into a coordinate system rotating with the rotor flux. The rotor flux and the corresponding angle are determined by a motor model. The control structure consists of a speed control with underlying torque /current control and flux control and works similar to a speed-controlled separately excited DC machine. In the field oriented coordinate system the motor current is considered as a vector consisting of two perpendular components:

• the longitudinal current generating the motor magnetisation (flux, excitation)
• the lateral current directly generating the motor torque (armature current)

The torque of the motor is proportional to the product of the magnetising and armature currents. The longitudinal current corresponds to the field current of a DC motor and the lateral current corresponds to the armature current. Whilst the field current has a long time constant and is therefore not suited to rapid response control, the armature current can be changed to directly correspond with the drive torque. For vector control the magnetising current is held constant, so that the motor always works with the same flux.The armature current is dependent on the motor loading. Vector control allows fast and immediate access to the motor torque and therefore a highly dynamic and robust control of the drive. If the speed of the induction machine exceeds the nominal speed, the inverter is no longer able to generate the voltage necessary to keep the flux constant. In this case, the flux is reduced such that the inverter output voltage is fixed at its maximum. This is known as the field weakening region with reduced motor torque.

 

In the case when an encoder is not available, speed and rotor position are calculated by the motor model, which is fed by the measured currents and voltages. For the motor model it is important that the motor parameters are set up correctly. The accuracy of the motor model and the motor parameters determine the accuracy of the speed control. A well adjusted system can achieve an accuracy significantly better than 5% of the nominal speed. If this is not sufficient for the application, an encoder must be used.

 

The signal processing required for the field orientated control of induction motors as well as the flux determination using a dynamic motor model is very complex. Although the theory was founded in the early seventies, practical applications could not be realized until cheap and powerful microprocessors became available. Today special microprocessors for drive applications have been developed, being able to perform all the signal processing for a high dynamic performance AC drive with a sampling period of about 50 microseconds.