The concept of Vector control

Hey Guys,
I am having a tough time in understanding the philosophy of vector control of induction motor.I have referred to atleast 10 - 15 papers on field oriented control on the internet but still at a loss in comprehending the concept in totality.
I fully understood the mathematics involved but not able to grasp the physical significance of the same & no books are offering the explanation.It is easy to state that FOC is similar to the control of DC motors but in reality it is difficult to visualise the several attributes.

Only two parameters can be changed to bring speed control in IMs

1.Input voltage
2.Input frequency

Using these how a control similar to DC motors is achieved is unclear to me!!

Can any one guide me some lucid explanation of vector control for my better understanding?..

I need to commission a drive which works on vector control next ,month & hence need clarity of the subject!!

pls help

With a DC brush motor, we vary voltage. For a stepper motor, the frequency.
It gets a bit more complex with AC induction motors.

1) To simplify, you first need to understand the concept of simple V/F control. A typical AC motor was designed to run at a fixed speed based on the AC frequency. Take a motor that's designed to run at a fixed speed at 60 Hertz. Typical is 1760 RPM. To change it's speed, we need to change the frequency of the AC we send to it. We also need to linearly reduce voltage with frequency to maintain torque. You said you understand the math. For more info, look at a Wiki for an induction motor.
This is called Open Loop control. We don't have an encoder, so we don't have any feedback.

2) An AC motor is an inductor, so it has a feedback signal. We use a fast microprocessor to "decode" this feedback and use it as our encoder.
It's also called Open Loop Control, because there's no physical encoder, but really it's closed loop, because we have a pseudo encoder.
This control method is called Open Loop Vector Control (OLV), or field-oriented control (FOC) . More info at this Wiki.

An article from Machine Design does a good job of putting these concepts in laymans terms.

keithkyll said:

With a DC brush motor, we vary voltage. For a stepper motor, the frequency.
It gets a bit more complex with AC induction motors.

1) To simplify, you first need to understand the concept of simple V/F control. A typical AC motor was designed to run at a fixed speed based on the AC frequency. Take a motor that's designed to run at a fixed speed at 60 Hertz. Typical is 1760 RPM. To change it's speed, we need to change the frequency of the AC we send to it. We also need to linearly reduce voltage with frequency to maintain torque. You said you understand the math. For more info, look at a Wiki for an induction motor.
This is called Open Loop control. We don't have an encoder, so we don't have any feedback.




2) An AC motor is an inductor, so it has a feedback signal. We use a fast microprocessor to "decode" this feedback and use it as our encoder.
It's also called Open Loop Control, because there's no physical encoder, but really it's closed loop, because we have a pseudo encoder.
This control method is called Open Loop Vector Control (OLV), or field-oriented control (FOC) . More info at this Wiki.

An article from Machine Design does a good job of putting these concepts in laymans terms.

Click to expand...

Thanks for the reply.I fully understand V/F control but when it comes to FOC the d axis currents & q axis current & its physical relevance is a concern in understanding the process

I found this from jraef. He is a frequent contributor on this forum. An excerpt for D and Q:
"It does this by keeping in mind that magnetizing current always lags (inductive) the voltage by 90 degrees and that the torque producing current is always in phase with the voltage. It controls the magnetizing current (usually named Id) in one control loop and the torque producing current (Iq) in another control loop. The two vectors Id and Iq, which are always 90 degrees apart, are then added (vector sum) and sent to the modulator, which turns the vector information into a rotating PWM modulated three-phase system with the correct frequency and voltage."

If you are just commissioning a drive, you really don't need to know theory, because the controller handles it all. Select your control method (OLVC), enter the motor parameters into the drive, then Autotune. The controller will measure resistance, inductance, etc. and store these parameters.

Originally posted bu rejoe.koshy:

It is easy to state that FOC is similar to the control of DC motors...

Click to expand...

That is something of an oversimplification. The only portion of the control law that would be similar to a DC motor are the regulation of Id and Iq, with Id being similar to field current control in a wound field DC motor and Iq similar to armature current control. This would allow you to use magnitude control techniques that act more similar to low frequency DC system than the higher frequency AC system being controlled. However, this is where the simiarity ends. How you get from motor current to Id_act and Iq_act and from Id_cmd and Iq_cmd to motor currents is what makes field oriented control possible.

I assume you have at least a conceptual feel for Clarke and Park transforms. If not, take a look at pages 7 and 8 of this:

http://www.microsemi.com/document-p...ations-mss-software-implementation-user-guide

Its pretty simplistic but gives you a feel for what is going on.

Originally posted by [/b]rejoe.koshy[/b]:

Only two parameters can be changed to bring speed control in IMs

1.Input voltage
2.Input frequency

Click to expand...

You are forgetting phase. The basis of FOC requires that Iq and Id remain at 90 degrees relative to each other. However, that whole reference frame can be rotated back and forth relative to the previous stator current commands. Just as importantly, changing the relative values of Id and Iq while holding the reference angle constant will result in both a change in the magnitude of the resulting current vector as well as the phase of the current vector. Generally speaking Id is held constant at motor speeds up to base speed, although it doesn't have to be. But simply changing Iq through its allowable range will often result in a change in the vector angle of the resultant of up to 45 degrees. This will have a direct result on the phase angle of the upcoming 3-phase motor voltage waveforms relative to the previous commanded waveforms.

Since this adjustment in the two current components and the resulting phase of the resultant current vector are typically not aligned with the Id axis for any appreciable period of time it looks like the frequency is being controlled directly to influence the torque in the motor. This is somewhat incidental to what is actually happening, though.

Keith

If you want a concept rather than the details, let's try this; it's part of a training curriculum I created for people who want to know, but don't need to make one from scratch...

With a V/Hz drive, you are sending out a speed command to the motor, but the drive actually has no idea what's going on in that motor (other than overload protection). In fact you can have the motor disconnected and a basic V/Hz drive would not even know. So if for example you send 30hz out of the drive to a 1780RPM motor, you are going to EXPECT that motor to spin at 890RPM (50% speed). If the disconnect swith is open and motor is dead, the drive has no idea, other than to tell you it is pulling no current.

But when connected to the motor, if the LOAD on the motor changes dramatically, the motor slows down, but the V/Hz drive still has no idea that it did. The motor slowing down will increase the slip, so the motor will pull more current and like all AC induction motors develop more torque to attempt to return to the normal slip speed. But the only thing the VFD sees is the increase in current, it has no idea why, nor does it react to that. It is "open loop".

With Vector Control, the loop is closed, in one way or another (more later). With a specialized microprocessor or DSP, the motor speed dropping under load is seen as an "error" in what boils down to a PID loop. To correct the error, the VFD alters the output to boost torque in the motor by boosting the current it draws. Within the way a motor draws current, the fist thing it does is use current to excite the steel and copper parts, called establishing the flux, then right after that to create magnetic field strength that creates torque in the motor. If you just boost the V/Hz ratio "willy-nilly", you end up boosting BOTH the flux producing current AND the torque producing current, even though increasing the flux current will saturate the windings and do no useful work. That means then that of the current you can send to the motor, a significant portion of it is not doing useful work, it's just heating it up unnecessarily. In the early days of VFDs (before vector control), we called this "torque boost", a setting you used to manually tweak the V/Hz ratio in order to increase the torque output. The drawback is/was, it simultaneously robbed you of capacity, so to avoid cooking your motor, you only used it at low speeds where the motor was (hopefully) less loaded.

Vector control is named for the fact that the flux current and torque current take place at different times, which can be represented by vectors. So vector control means the drive's processor is tracking that error and boosting the torque vector WITHOUT boosting the flux vector any more than is necessary. Doing so means more (most) of the available current the motor can handle is realized as torque to correct that error in the motor speed without wasting any of it as useless and destructive excess flux. That extra torque also means that the motor can respond to and correct that error extremely fast without very much overshoot, so your speed regulation improves dramatically. In V/Hz your ability to regulate speed, even on a stead state load, is good for about +-1% (depending on drive quality). With even simple Sensorless Vector Control that becomes 0.1%, and with FOC is can be as high as 0.001%, because the motor can respond so quickly to an error.

Sensorless Vector Control is not really sensorless, the sensors are in the VFD, looking at the current waveform going to the motor and tracking anomalies that indicate the rotor moving through the stator fields. So it uses that relative position response to correct the error. In that process, it is not fully separating the vectors so much as estimating the flux vector to maximize the torque vector, based on a mathematical model of that motor circuit you created inside of the drive via testing.

Closed Loop Flux Vector Control and FOC generally use an encoder on the motor to feed an ABSOLUTE position of the rotor back to the microprocessor, so that the drive can fully separate the flux and torque vectors at any moment in time and tweak that response even tighter. To do torque control instead of just speed control, it has to have separate regulator loops for the vectors, so it means a more powerful microprocessor, which is what you have in those drives. With that, you attain the "holy grail" of Vector Control AC drives; the ability to produce 100% torque at zero speed, something you must have when doing hoist control. By totally controlling the vectors separately, you can (for a short time) flux the motor exactly as much as it needs without cooking it while making the shaft hold its position without moving when you let go of a mechanical brake. Some newer FOC drives can now even accomplish this without the external encoder as well.

FVC/FOC using the encoder can also make a standard AC induction motor perform almost as well as a servo (I said ALMOST). Although not something you would want to use to replace a servo, if you needed a 75HP servo, you are not going to find one. This creates the next best thing, which in most applications needing high HP is going to be the best you can get.

Somewhere in between SVC and FVC is "Velocity Vector Control" that uses an encoder for feedback to improve speed accuracy over SVC, but still has the single regulator loop assigned only to speed (velocity) errors. You still get BETTER torque capability, it's just not totally in control of torque like FOC.

jraef said:

If you want a concept rather than the details, let's try this; it's part of a training curriculum I created for people who want to know, but don't need to make one from scratch...

Click to expand...

Thank you so much!... that clears a lot of confusion but still have a few doubts.

I did the math in an excel sheet to see the effect on terminal voltages imposed on the motor by changing the I d & I q current values.The results show me that the voltages increase with the increase when I increase I d or i q values.It also changes the position of the resultant current vector in the motor.This means that I can get more torque by varying of the quantity (I d or I q) ie either of the quantities.But according to the concept introduced by you, the torque producing part is I q only.....

One thing to keep in mind is that most of the FOC drives will keep Id constant. The motor will be somewhat more responsive if you keep it fully fluxed and vary only the torque producing current component. It doesn't necessarily have to be this way but it is with many drives.

A motor will not produce any torque if all the motor current is on the Id vector. The motor is basically acting like an inductor and the phase currents are being controlled so they produce little or no current in the rotor. As torque is required and the Iq vector grows it has the dual effect of increasing the overall motor resultant vector length, which increases total motor current, as well as rotating the resultant vector off of the Id axis, which has the effect of producing a phase current sequence that will result in induced rotor current and, therefore, torque production. If all you were looking at was phase voltage and frequency the added torque requirement would probably show itself as a slight increase in frequency if the torque requirement stayed steady.

Keith

Just wanted open up the discussion on vector control.After browsing through a lot of literature on vector control, i am finally beginning to understand the concept a little bit.However, every explanation is based on the fact that the motor is star connected.
But in reality , we have a lot of delta connected motors.Wondering how FOC would work for a delta connected motor, since the entire algorithm depends on the measured line currents, which in in case of star connection will be equal to phase current but does not hold true for delta connection.

How does FOC work for delta connected motors??

Whether the motor is designed as star or delta windings for the voltage you are using has no bearing on performance one way or the other.

jraef said:

Whether the motor is designed as star or delta windings for the voltage you are using has no bearing on performance one way or the other.

Click to expand...

But the feedback is phase currents so in case of delta winding, how will drive measure the phase currents??

You are confusing how the current us used in the motor with the how that current flows out of the drive. You have 3 phases leaving the drive, the drive is monitoring those 3 phases. Don't over think it.