PID Controller and Tuning Flashcards

1
Q

What is a P controller?

A

It is essentially an amplifier with an adjustable gain

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

What happens if we adjust Kp?

A

it will adjust the rise time and steady state error. This can be unwanted in higher order systems as it can give unwanted oscillation.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

What is the equation for a PI controller?

A

Kp + Ki / s

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

What is the type number of a PI controller?

A

Type 1

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

What is the steady state error of a PI controller to a Step Input?

A

0 because it is a type 1 system.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

Why does a PI controller have the potential to be less stable?

A

Because it increases the order of the system.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

What is the equation for a PD controller?

A

Kp + Kds.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

What does the derivative term add to the system?

A

It adds damping.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

What effect does damping have on the system?

A

It can reduce overshoot and rise time along with stabilise an unstable system.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Why is a PD controller hard to implement in real life?

A

Noise.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

What is the equation for a PID controller?

A

Kds^2 + Kps + Ki / s

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

What is the effect on the rise time, overshoot, settling time and steady state error of the system if we increase Kp?

A

Rise Time: decrease
Overshoot: increase
Settling Time: small change
Steady State Error: decrease

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

What is the effect on the rise time, overshoot, settling time and steady state error of the system if we increase Ki?

A

Rise Time: decrease
Overshoot: increase
Settling Time: increase
Steady State Error: ELIMINATES

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

What is the effect on the rise time, overshoot, settling time and steady state error of the system if we increase Kd?

A

Rise Time: small change
Overshoot: decrease
Settling Time: decrease
Steady State Error: small change

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

What is good about the Ziegler Nichols Tuning Method?

A

We don’t need to know the model of the system.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

What is the transport delay in the Ziegler Nichols Method?

A

It is the time taken for the signal provided to be acted upon ( the time taken to see a change in the system ).

17
Q

What is a model of the Open Loop ZN tuning method?

A

G(s) = Ke^-Td*s / Tau * s + 1

18
Q

What is the Open Loop ZN model similar to?

What effect does multiplying by the exponential have?

A

The time constant form for a first order system.

Multiplying by the exponential moves everything to the right.

19
Q

What is the ZN gain Kp for a P controller?

A

Tau / K * Td

20
Q

What are the ZN gains Kp and KI for a PI controller?

A
Kp = 0.9 * Tau / KTd
Ki = 0.27 * Tau / KTd^2
21
Q

What are the gains for the ZN PID controller?

A
Kp = 1.2Tau / KTd
Ki = 0.6Tau / KTd^2
Kd = 0.6Tau / KTd
22
Q

What are the necessary conditions to use the open loop tuning method?

A

The system must be stable and have an S shape with delay.

23
Q

For the ZN closed loop tuning method, what must the state of the system be?

A

It can be stable or unstable, it doesn’t matter since we don’t need a model of the system.

24
Q

How do we apply the ZN closed loop tuning method?

A

Start with just P control. Increase the gain until we get sustained oscillations. This means that it is marginally stable. We record this gain as the ULTIMATE GAIN (Ku). We take the period of Oscillation (Tu) as the ULTIMATE PERIOD. From these two values we can work out any P, PI, or PID controller gains.

25
Q

What is the ULTIMATE GAIN?

A

It is the gain at which the system will be marginally stable and have sustained oscillations.

26
Q

What is the ULTIMATE PERIOD?

A

It is the period of oscillations at the ULTIMATE GAIN.

27
Q

What is the gain Kp for a P controller using the ZN CL method?

A

0.5*Ku

28
Q

What is the gain Kp and KI for a PI controller using the ZN CL method?

A
Kp = 0.45*Ku
Ki = 0.54*Ku / Tu
29
Q

What is the gain Kp, Ki, and Kd for a PID controller using the ZN CL method?

A
Kp = 0.6*Ku
Ki = 1.2*Ku / Tu
Kd = 0.075Ku*Tu.
30
Q

Why can the ZN CL method be bad?

A

Because at sustained oscillation the system may get damaged.

31
Q

Why is PID bad for a noisy system?

A

The Derivative term will only amplify that noise and can lead to unwanted variations in the system input.

32
Q

How do we mitigate the risk of the Pd term amplifying the noise?

A

Pass it through a low pass filter or don’t use the Pd controller at all.

33
Q

What is the problem with filtering the Kd term?

A

Although it may smooth out the signal, the filter may add a delay which can affect the results.

34
Q

What is integral windup?

A

If the input is saturated then the integral term can grow very large before reaching its reference value and reducing again. This will lead to a very large overshoot.

35
Q

How do we mitigate integral windup?

A

Put a limit on how big the integral term can grow.

36
Q

What is gain scheduling?

A

It is used for a non linear system where the gains do not act around a single point. It will change the gains depending on the range of operation.