PID Controller Design,
Chapter 12
Tuning and Troubleshooting
Desirable Controller Features
Chapter 12
0. Stable
1. Quick responding
2. Adequate disturbance rejection
3. Insensitive to model, measurement errors
4. Avoids excessive controller action
5. Suitable over a wide range of operating conditions
Impossible to satisfy all 5 unless self-tuning.
Alternatives for Controller Design
[Link] correlations most limited to 1st order plus
dead time
Chapter 12
[Link]-loop transfer function - analysis of stability or
response characteristics.
[Link] simulation (requires computer software
like MATLAB and Simulink)
[Link] response - stability and performance
(requires computer simulation and graphics)
[Link]-line controller cycling (field tuning)
Controller Synthesis - Time Domain
Time-domain techniques can be classified into two
groups:
(a) Criteria based on a few points in the response
Chapter 12
(b) Criteria based on the entire response, or integral
criteria
Approach (a): settling time, % overshoot, rise time, decay
ratio
Ke s
Process model G( s) (1st order)
s 1
Several methods based on 1/4 decay ratio have been
proposed: Cohen-Coon, Ziegler-Nichols
Approach (b)
1. Integral ofsquare error (ISE)
ISE e( t ) dt
2
0
2. Integral of absolute value of error (IAE)
IAE e( t ) dt
Chapter 12
0
3. Time-weighted IAE
ITAE t e( t ) dt
0
Pick controller parameters to minimize integral.
ISE penalizes large errors, produces the most aggressive tunings
IAE allows larger deviation than ISE (smaller overshoots)
ITAE penalizes errors that persist over longer periods, produces the most conservative
controller tunings
Approximate optimum tuning parameters are correlated
with K, ,
ITAE tunings assuming a FOPDT processa
Chapter 12
Tuning Relationships
1. KC is inversely proportional to KPKVKM .
Chapter 12
2. KC decreases as / increases.
3. I and D increase as / increases (typically D =
0.25 I ).
4. Reduce Kc, when adding more integral action.
5. To reduce oscillation, decrease KC and increase I .
Disadvantages of Tuning Correlations
1. Stability margin is not quantified.
Chapter 12
2. Control laws can be vendor - specific.
3. First order + time delay model can be inaccurate.
4. Kp, , and can vary.
5. Resolution, measurement errors decrease stability
margins.
6. decay ratio not conservative standard (too
oscillatory).
Direct Synthesis Method
Y GC G
( G includes Gm, Gv and Gp)
Ysp 1 GC G
Chapter 12
1. Specify DESIRED closed-loop response (in transfer function form)
Y
Gdesired (s)
Y
sp d
2. Get process model, G (= GPGMGV)
3. Solve for Gc,
Y
1 Ysp d
GC
G
1 Y
Ysp
d
Specify desired Closed Loop Response
Y Y
Define, (s ) 1
Gdesired
Y Y s 1
d
sp sp d c
Chapter 12
(first order response, no offset)
c desired speed of response
1 1
Yields Gc ~
G cs
K s +1 1
For G = , Gc = = + (PI)
s +1 K cs K c K cs
Derivation of PI Controller for FOPTD Process
Consider the standard first-order-plus-time-delay model,
Y
Gdesired (s)
Chapter 12
Y
sp d
Solve for the closed-loop response: Y
1 Ysp d
GC
G
1 Y
Ysp
d
Then approximate e- s - 1 - s.
1 e s
Gives, Gc ~
G ( c ) s
Derivation of PI Controller for FOPTD Process
1 e s
Substituting in Gc ~
G ( c ) s
Chapter 12
and rearranging gives a PI controller with,
Gc Kc 1 1/ I s ,
Second-Order-plus-Time-Delay (SOPTD) Model
Consider a second-order-plus-time-delay model,
1 e s
Chapter 12
Use of FOPTD closed-loop response and time delay Gc ~
G ( c ) s
gives a PID controller in parallel form,
where
Example
Use the DS design method to calculate PID controller settings for
the process:
2e s
G
10s 1 5s 1
Chapter 12
Consider three values of the desired closed-loop time constant:
c = 1, 3, and 10. Evaluate the controllers for unit step changes in
both the set point and the disturbance, assuming that Gd = G.
Perform the evaluation for two cases:
a. The process model is perfect ( G = G).
b. The model gain is K = 0.9, instead of the actual value, K = 2.
This model error could cause a robustness problem in the
controller for K = 2.
0.9e s
G
10s 1 5s 1
The controller settings for this example are:
c 1 c 3 c 10
Kc K 2 3.75 1.88 0.682
Kc K 0.9 8.33 4.17 1.51
I 15 15 15
3.33 3.33 3.33
D
Note only Kc is affected by the change in process gain.
The values of Kc decrease as c increases, but the values of I
and D do not change
Chapter 12
Correct model gain
Chapter 12
Incorrect model gain.
Controller Tuning Relations
Model-based design methods such as DS produce PI or PID
controllers for certain classes of process models, with one tuning
parameter c
Chapter 12
How to Select c?
Several guidelines for c have been published for FOPDT
models:
1. c / > 0.8 and c 0.1 (Rivera et al., 1986)
2. c (Chien and Fruehauf, 1990)
3. c (Skogestad, 2003)
For example by setting c / 3 , the desired closed loop response
is three times faster than the open-loop response
On-Line Controller Tuning
1. Controller tuning inevitably involves a tradeoff between
performance and robustness.
2. Controller settings do not have to be precisely determined. In
Chapter 12
general, a small change in a controller setting from its best
value (for example, 10%) has little effect on closed-loop
responses.
3. For most plants, it is not feasible to manually tune each
controller. Tuning is usually done by a control specialist
(engineer or technician) or by a plant operator. Because each
person is typically responsible for 300 to 1000 control loops, it
is not feasible to tune every controller.
4. Diagnostic techniques for monitoring control system
performance are available.