Faculty of Engineering

Mechanical Engineering Department

Industrial Process Control

(22MTRN12I)
(Lect. 3)
Dr. Nader A. Mansour
[email protected]
Previous Lecture

PID Controller
Previous Lecture

 Proportional controller (P)

reduces error responses to disturbances,
speeds up the process response
but still allows a steady-state error.
 Integral controller (I)
The the steady state error to a constant input can be
eliminated, at the cost of deterioration in the dynamic
response.
 Derivative controller (D)
typically makes the system better damped and more
stable.
Integral Windup

Due to limits of max. input accepted by process

Or max. output that can be created by the actuator.

 uin , if , uin  u0

1 1
uout =  + u0 , if , uin  u0
u_In u_Out
− u , if , u  −u
Saturation  0 in 0
Integral Windup

Effect of Saturation
Damage to the actuator due to the excessive input signals.
Slow down the system behavior.
Nonlinear behavior of the control signal.
Integral Windup

Simulink Model
Anti-Windup Mechanisms

Integrator clamping (conditional integrator)

Stops the integrator term during saturation
Back calculation
Generates feedback signal to modify the integrator value.
Integrator Clamping

Integrator is switched off upon saturation

Integrator Clamping

Simulink Model
Back Calculation

Feedback signal is generated only when the actuator is

saturated
Back Calculation

Simulink Model
PID Tuning

How to find the controller gains??

Analytical Tuning Techniques
Frequency response analysis
Empirical Tuning Techniques
On-line tuning after the control system is installed
Computer simulation
PID Tuning

Empirical Tuning Techniques

Two Steps:
Determination of the dynamic
characteristics of the control loop
Estimation of the controller tuning
parameters that produce a desired
response
Tuning Techniques
Ziegler-Nichols Oscillation Method
Ziegler-Nichols Oscillation Method

Is also called Ziegler-Nichols closed loop

method.
It uses the proportional controller to force
sustained oscillations.
The goal is to determine ultimate gain and
ultimate period of oscillation of the control loop
Ziegler-Nichols Oscillation Method

Ziegler-Nichols closed loop Ziegler-Nichols

method specifies a decay ratio of one-fourth
(1/4)
Decay ratio = ratio of the amplitudes of two
successive oscillations
Ziegler-Nichols Oscillation Method

Set the true plant under proportional control,

with a very small gain
Increase the gain until the loop starts oscillating
Note that linear oscillation is required and that it
should be detected at the controller output.
Record the controller critical gain (ultimate
gain) Kp >> Kc or Ku and the oscillation period
of the controller output, Pc or Pu.
Ziegler-Nichols Oscillation Method

Calculate the parameters according to the

following table Controller Proportional Integral Derivative
Type Gain (Kp) Time (tI) Time (tD)
P-only Ku/2 - -
PI Ku/2.2 Pu/1.2 -
PID Ku/1.7 Pu/2 Pu/8
Ziegler-Nichols Oscillation Method

Drawbacks
Time-consuming
If the process dynamics are slow
The process is pushed to the stability
limits
Risk of plant damage
Ziegler-Nichols Oscillation Method

Controller Proportional Integral Derivative

Type Gain (Kp) Time (tI) Time (tD)
P-only Ku/2 - -
PI Ku/2.2 Pu/1.2 -
PID Ku/1.7 Pu/2 Pu/8
Assuming unity feedback system with controller,

𝐺𝐶 = 𝐾
Now the CLTF will be as follows;
𝐺𝑃 𝐺𝑐 𝐾
𝐺 𝑠 = = 3
1 + 𝐺𝑃 𝐺𝐶 𝑆 + 6.2𝑆 2 + 6.2𝑆 + 1 + 𝑘
Routh-Hurwitz Criterion should be used to find
the stability margin of the CLTF as follows;

𝑺𝟑 1 6.2
Then, the Ultimate gain
𝑺𝟐 6.2 1+K Ku = 37.44
𝑺𝟏 (37.44−K)
𝟔. 𝟐
𝑺𝟎 1+K
 The period of oscillation at this ultimate gain must be calculated
by the direct substitution s = ωj into the characteristic equation
as follows;
−𝑗𝜔3 − 6.2𝜔2 + 6.2𝑗𝜔 + 1 + 𝑘 = 0
1 + 𝑘 − 6.2𝜔2 + 𝑗 6.2𝜔 − 𝜔3 = 0

 Setting the imaginary part to zero gives the frequency of

oscillation and the Ultimate Period Pu,
𝜔 = 2.48 𝑟𝑎𝑑/𝑠𝑒𝑐
2𝜋
𝑃𝑢 = = 2.52 𝑠𝑒𝑐
𝜔
From the ZN table, 𝐾𝑝 22.02
Kp= Ku/1.7 = 37.44/1.7 = 22.02 𝐾𝑖 = = = 17.47
𝑇𝑖 1.26
Ti= Pu/2 = 2.52/2 = 1.26
Td= Pu/8 = 2.52/8 = 0.315 𝐾𝑑 = 𝐾𝑝 * 𝑇𝑑 = 22.02 ∗ 0.315 = 6.93
Analog PID Controller
 Analog PID Controller
Build the
following
circuit of the
PID Controller
using the
values found
in the next
table.
Analog PID Controller
Analog PID Controller
Questions??!