Professional Documents
Culture Documents
In this section, we analyze PD and PID control of a plant typical in mechanical positioning systems. We also propose a possible design method. The nominal model for the plant is A P (s) = s(s + p) where A and p are xed parameters.
13.1
PD control
First, consider PD control, specically proportional control, with inner loop ratefeedback. This is shown below (its just the PID diagram, with the integral action removed) d r
E i EK P E i E
c E A E i E
1 s+p
E c
1 s
E c
KD ' '
'
Inner-loop Rate-Feedback
In terms of plant and controller parameters, the loop gain (at breaking point marked by ) is A(KD s + KP ) L(s) = s(s + p) In other words, from a stability point of view, the system is just unity-gain, negative feedback around L.
E KP
E i
c E A E i E
1 s+p
E c
1 s
E c
KD ' '
'
126
The closed-loop transfer function from R and D to Y is Y (s) = s2 AKP 1 R(s) + 2 D(s) + (AKD + p)s + AKP s + (AKD + p)s + AKP
Clearly, with two controller parameters, and a 2nd order closed-loop system, the poles can be freely assigned. Using the (, n ) parametrization, we set the characteristic equation to be 2 s2 + 2n s + n giving design equations
2 n , KP := A
KD :=
2n p A
In terms of the (, n ) parametrization, the loop gain and transfer functions are L(s) = Y (s) =
2 (2n p)s + n s(s + p)
2 1 n R(s) + 2 D(s) 2 + 2 + 2 2 s s + 2n + n n n
(66)
Although this is a 2nd order system, and most quantities can be computed analytically, the formulae that arise are rather messy, and interpretation ends up requiring plotting. Hence, we skip the analytic calculations, and simply numerically compute and plot interesting properties for dierent values of n , p and . Normalization is the key to displaying the data in a cohesive and minimal fashion. For now, take p = 0 (you should take the time to write a MatLab script le that duplicates these results for arbitrary p). In this case, it is possible to write everything in terms of normalized frequency, all relative to n . This simultaneously leads to a normalization in time (recall homework 8). Hence frequency responses are plotted G(j) versus n , and time responses plotted y(t) versus n t. We consider a few typical values for .
127
The plots below are: Magnitude/Phase plots of Loop transfer function. These are normalized in frequency, and show L(j) versus n . From these, we can read o the crossover frequencies and margins.
OpenLoop Transfer Function, PD
10
10
10
xi = 1.3
Magnitude
10
10
10
10
10
10
10
10
Phase (degrees)
130 140 150 160 170 180 2 10 xi = 0.5 xi = 0.707 xi = 0.95 xi = 1.3
10
10
10
128
Magnitude/Phase plots of closed-loop R Y transfer function. These are normalized in frequency, and show GRY (j) versus n .
R>Y Frequency Response, PD
10
10
10
Magnitude
10
10
xi = 1.3
10
10
10
10
10
Phase (degrees)
10
10
10
129
Magnitude plot of closed-loop D Y These are normalized in frequency 2 and magnitude,, and show n GDY (j) versus n .
Normalized Disturbance Response, PD
10
xi = 0.5 xi = 0.707 10
0
xi = 0.95 xi = 1.3
10
10
10
10
10
10
10
10
Unit step d y responses. These are normalized both in time, and in 2 response. Hence the plot is n y(t) versus n t.
0.8
0.6
130
Unit step r y responses. These are normalized in time, and show y(t) versus n t.
0.8
0.6
Magnitude plot of closed-loop R E. These are normalized in frequency, and show GRE (j) versus n .
R>E (Sensitivity), PD
10
10
Magnitude
xi = 0.5 10
1
10
10
10
10
10
The r y response has the canonical 2nd order response we have come to know and love. The steady-state disturbance rejection properties are dependent on n . As n increases, the eect of a disturbance d on the output y is decreased. Hence, in order to improve the disturbance rejection characteristics, we need to pick larger n . Depending on , the gain-crossover frequency is between about 1.3n and 2.5n . So, using this controller architecture, the gain crossover frequency must increase when the steady-state disturbance rejection is improved. The phase margin varies between 53 and 83 . There is no phase-crossover frequency, so as dened, the gain margin is innite. The closer that the complex frequency response remains to 1 (over a large frequency range), the better the r y response. The term bandwidth is often used to mean the largest frequency B such that for all satisfying 0 B , |1 GRY (j)| 0.3 Be careful with the word bandwidth. Make sure whoever you are talking to agrees on exactly what you both mean. Sometimes people use it to mean the gain crossover frequency. Generally, the higher the bandwidth, the faster the response, and better the disturbance rejection. Of course, its hard to explicitly assess time-domain properties from a single number about a frequency response, so use it carefully. The same types of intuition can also be assessed by looking at the frequency range over which the transfer function is GRE small, and also verifying that it is not too large in another range.
13.2
PID Control
In order to reduce the steady-state eect of the disturbance, we next analyze PID control, namely proportional+integral, with inner loop rate feedback. This is shown below. d r
E i E E KI E i T ! E KP E
c E A E i E
1 s+p
E c
1 s
E c
KD ' '
'
Inner-loop Rate-Feedback
132
The closed-loop transfer function is Y (s) = s A(KP s + KI ) R(s) + 3 D(s) s3 + (p + AKD )s2 + AKP s + AKI s + (p + AKD )s2 + AKP s + AKI
The closed-loop characteristic equation is s3 + (p + AKD )s2 + AKP s + AKI With three controller parameters, and a 3rd order closed-loop system, the poles can be freely assigned. Using the (, n ) parametrization, along with a 3rd pole at n , we set the characteristic equation to be CE : Multiplied out, this gives
2 3 s3 + (2 + )n s2 + (2 + 1)n s + n 2 (s2 + 2n s + n )(s + n )
Choosing specic values of , n and yields appropriate controller gains, via the design equations, which are obtained by simply equating coecients, KD = (2 + )n p , A KP =
2 (2 + 1)n , A
KI =
3 n A
Note that for 1, the design equations give KD and KP as in the PD case, along with a very small integral control term. Hence, for a given pair (, n ), picking small and doing the full PID design is equivalent to doing the PD design for and n , and then simply adding a small amount of integral control as an afterthought. That approach will leave a closed-loop pole near the origin, approximately at s = AKI (= n ). 2
n
The steady-state gain from d to y is zero, due to the integral term. Again, take the case p = 0. For clarity, lets also pick = 0.707, and only study the variation in responses due to our choice of . Again, the normalization with n is complete, in both time and frequency, with frequency responses plotted versus n , and time responses plotted versus n t.
133
alpha = 0.1
5
Magnitude
10 10 10 10 10 10
10
10
10
10
250
Phase (degrees)
200
150
100
50 2 10
10
10
10
134
10
10
Magnitude
alpha = 0.1 10
2
10
alpha = 10
10
10
10
10
10
Phase (degrees)
10
10
10
135
10
| wn^2 G |
10
10
10
10
10
10
0.8
wn^2 y
0.6
0.4
0.2
0.2 0
136
1.2
0.8 alpha = 0.1 0.6 alpha = 0.3162 alpha = 1 0.4 alpha = 3.162 alpha = 10 0.2
Y
0 0
R E magnitude plots
10
1
10
10
Magnitude
10
10
10
10
10
10
10
For a given n and > 1, both the crossover frequencies, and bandwidth (look at GRE ) are much higher than the PD case. This is somewhat reected in quicker rise times and comparable settling times. The gain crossover frequency increases signicantly with increasing . For instance, at 0 (which is the same as the PD control) the crossover frequency is about 1.7n , whereas for the crossover frequency jumps to approximately 5n at 3.1. At the respective crossover frequencies, the phase margins of eth PD and PID designs are similar. As increases, the disturbance rejection properties change. Any (and every) > 0 gives perfect steady-state disturbance rejection, but the time-domain and frequency domain properties for dierent are quite dierent. It is instructive to calculate the residue associated with the pole at n when r(t) is a unit step. It is then fairly easy to explain the slow settling times that occur for the intermediate values of . Remember, in typical applications, the uncertainty in the plants behavior increases with increasing frequency, so designs that lead to higher crossover frequencies usually are required (for condence) to have signicantly larger phase margins. Usually, for a given problem, modeling innaccuracies and unknown dynamics typically impose a maximum allowable crossover frequency, regardless of phase margin.
138
Since some normalization is possible (using n ), brute-force repeated simulation allows us to approximately compute several functions. They are functions of p, and . Here, we imagine that p is known, and xed. We also propose to x = 0.707, leaving only functions of . In any given design situation, it may be necessary to modify the choice of , and recompute. The functions are plotted below. normalized crossover frequency versus
100
80
60
wc/wn
40 20 0 2 10
10
10 Alpha
10
10
139
85
80
PHI
75
70
65
60 2 10
10
10 Alpha
10
10
140
2.5
wn*Trise
2 1.5 1 2 10
10
10 Alpha
10
10
20
wn*Tsettle
15
10
0 2 10
10
10 Alpha
10
10
141
0.8
wn^2 y
0.6
0.4
0.2
0 2 10
10
10 Alpha
10
10
350
300
250
wn*Tsettle
200
150
100
50
0 2 10
10
10 Alpha
10
10
142
30
25
% Overshoot
20
15
10
5 2 10
10
10 Alpha
10
10
Suppose that these have been computed reasonably accurately, at suciently large numbers of values. Can we use all of this data to develop a foolproof design method?
13.3
In designing the PID controller gains, the free parameters (at this point) to be chosen are , n and . For xed choice of , we can precompute functions f1 , f2 , . . . , f5 of such that 1. Gain crossover frequency (c ) equals n f1 () 2. Rise Time (tR ) equals f2 ()/n 3. Settling Time (tS ) equals f3 ()/n
2 4. Peak response to step disturbance (yd,max ) equals f4 ()/n
5. Settling time of step disturbance response (tS,d ) equals f5 ()/n So, given target requirements, we can fairly easily determine if there is a PID controller which satises the objectives. Specically, take objectives as c c , tR tR , tS tS , yd,max yd,max , tS,d tS,d 143
where the over-bar quantities are targets. Hence we search for values of n and which satisfy
f () f () 2 3 gL () := max , , tR tS
Hence, we simply graph the two functions gL and gU , and see if there is any value of where gl () gu (). If so, then simply pick an for which the inequality it true, and pick any n such that
gL ( ) n gU ( )
f4 () f5 () c , =: gU () n yd,max tS,d f1 ()
Moreover, since the overshoot reaches a maximum at = 1, and falls o on both sides, you can easily reduce the overshoot by moving to one side of the feasible region.
13.4
Design Example
Consider the Lab, with single inertia, and pulley. The PD control worked reasonably well with n = 25, and = 0.707. This implies that a phase margin of 65 at a crossover frequency of 38 is adequate for stability robustness. So, in designing a PID controller, lets aim for a crossover frequency of 38, a rise time of 0.4 seconds, settling time of 0.55 seconds, and a disturbance response settling time of 0.7 seconds. Well set the peak disturbance response at 5, which essentially makes it not relevant, and then we could tighten down on it if we wanted. The constraints on and n are shown below
10
3
10
10
wn
10
0
10
10
10
10
10 Alpha
10
10
144
The circle is the point I chose, which MatLab tells me is = 0.36, n = 19.2
which is pretty similar to what we had working in the lab. Plots of the various relevant quantities Open-Loop gain
10
2
OpenLoop Gain
10
| L(j w) |
10
10
10
10
10
10 Frequency, rad/sec
10
145
Open-Loop Phase
OpenLoop Phase 280 260 240 220 200 180 160 140 120 100 0 10
10
10 Frequency, rad/sec
10
1.2
0.8
y
0.6 0.4 0.2 0 0
0.1
0.2
0.3
0.4
0.7
0.8
0.9
146
1.4
1.2
0.8
0.6
0.4
0.2
0 0
0.1
0.2
0.3
0.4
0.7
0.8
0.9
147