3) Safety Instructions
3.1 Danger to humans
- DANGER! Exercise caution when handling electrical system components. There is a danger of electric shock. Disconnect the mains plug before accessing any electrical components. Such work should only be performed by qualities personnel.
- DANGER! Never operate the device without a correctly installed ground conductor. A failure to observe this instruction might result in harm to humans and equipment.
3.2 Danger to Equipment and Functionality
- Caution! Do not fill the water tank with more than 15 litres of water. Excess water might overflow into the device and damage it.
- Caution! Never operate the pump without water. Dry running can damage the pump.
- Caution! Do not change the sensor's basic setting. Altered signals can result in a loss of process control.
- Caution! Drain the water tank prior to shutdown periods of more than 3 weeks.
- Caution! Store the model under frost-free conditions. Frost can damage individual components.
4) Experiments
4.1 Response of the control system
- This experiment is intended to ascertain the properties and response of the control system underlying the model.
- For this purpose, steps in control value are applied successively to the system in the non-regulated mode and the system's response is observed.
Experiment procedure:
- Turn on the demonstration model via its main switch.
- Fully open the adjustment cock (6). Setting: 0 degree.
- Set the controller to manual operation and the manipulated variable y to 10% =(2.5 litre/min).
- Turn on the pump. After a certain time, the flow rate assumes a constant value. Read and note this value.
- Increment the manipulated variable successively by 10%, waiting briefly each time until the flow rate has attained a constant value. Read and note these values too.
Manipulated variable y in %
|
10
|
20
|
30
|
40
|
50
|
60
|
70
|
80
|
90
|
100
|
Flow rate in (litre/min)
|
0.25
|
0.27
|
0.51
|
0.78
|
1.47
|
2.0
|
3.5
|
5.8
|
7.1
|
7.65
|
Result:
- The system evidently responds very quickly to change in the valve setting, much faster than temperature, filling-level and pressure control systems.
- The system is of a compensatory nature, resulting in constant final values each time.
- This characteristics was expected, because every pipe system possesses an intrinsic resistance to flow, thus preventing flow rates from rising indefinitely.
- The characteristic of the control variable x clearly indicates the equal-percentage response of the control valve.
4.2 Flow Control with a PI Controller
- In this experiment, a controller with proportional and integral components is used for flow control, accompanied by variations in parameters.
- The controller's differential component remains inactive.
- The control circuit's response to changes in the reference variable w is observed.
4.2.1 Slow PI-Controller
Experiment procedure:
- Turn on the demonstration model via its main switch.
- Set the controller and demonstration model as shown in the following table:
Controller type
|
PI-Controller
|
Controller mode
|
Automatic
|
P-component
>> Pb.1
|
0.1
|
I-component >> rt
|
4 seconds
|
D-component
>> dt
|
0.0 second
|
Controller settings
Start value:
Step
value:
|
6 litre/min (30%)
12 litre/min (60%)
|
Adjustment cock
|
Half open, 45 degree
|
- Observe the flow rate using the readings indicated by the controller and rotameter. After a certain time, the flow rate assumes a constant value of 6 litre/min.
- Increment the reference variable w by setting the controller to 12 litre/min. The flow rate increases and assumes a constant value of 12 litre/min after a certain time.
Result:
- The input signal y reveals that although the controller responds immediately to changes in the reference variable, it takes a long time to achieve a constant target value. The desired flow rate of 12 litre/min is attained very slowly ( t > 1 minute )
- The controller's P-component achieves fast response, while the I-component eliminates persistent deviations. However, the selected integration time is still too long.
4.2.2 Fast PI-Controller
- In this experiment,too, a controller with proportional and integral components is used for flow control.
- The controller's differential components remains inactive.
- Compared with experiment 4.2.1, the integration time.- i.e. the controller's I-component is set to a notably lower value.
- The control circuit's response to changes in the reference variable w is observed.
Experiment procedure:
- Turn on the demonstration model via its main switch.
- Set the controller and demonstration model as shown in the following table:
Controller type
|
PI-Controller
|
Controller mode
|
Automatic
|
P-component
>> Pb.1
|
0.1
|
I-component >> rt
|
0.5 seconds
|
D-component
>> dt
|
0.0 second
|
Controller settings Start value: Step
value:
|
6 litre/min (30%)
12 litre/min (60%)
|
Adjustment cock
|
Half open, 45 degree
|
- Observe the flow rate using the readings indicated by the controller and rotameter. After a certain time, the flow rate assumes a constant value of 6 litre/min.
- Increment the reference variable w by setting the controller to 12 litre/min. The flow rate increases and assumes a constant value of 12 litre/min after a certain time.
Result:
- The input signal y reveals that the controller quickly generates values which are notably higher than in the previous experiment. In fact, the control variable now distinctly overshoots the target value of 12 litre/min and starts to oscillate about it.
- The oscillations decay in 30 seconds to a permanent level of roughly +/- 5% about the target value.
- The parameters selected here do not result in satisfactory control performance. The selected integration time is obviously too short.
4.2.3 PI-Controller with Improved Parameters
- In this experiment, too, a controller with proportional and integral components is used for flow control.
- The controller's differential component remains inactive.
- The results obtained in experiments 4.2.1 and 4.2.2 are used as a basis for adapting the controller's integration time here.
- The control circuit's response to changes in the reference variable w is observed.
Experiment procedure:
- Turn on the demonstration model via its main switch.
- Set the controller and demonstration model as shown in the following table:
Controller type
|
PI-Controller
|
Controller mode
|
Automatic
|
P-component
>> Pb.1
|
0.1
|
I-component >> rt
|
0.75 seconds
|
D-component
>> dt
|
0.0 second
|
Controller settings
Start value:
Step
value:
|
6 litre/min (30%)
12 litre/min (60%)
|
Adjustment cock
|
Half open, 45 degree
|
- Observe the flow rate using the readings indicated by the controller and rotameter. After a certain time, the flow rate assumes a constant value of 6 litre/min.
- Increment the reference variable w by setting the controller to 12 litre/min. The flow rate increases and assumes a constant value of 12 litre/min after a certain time.
Result:
- The input signal y rises immediately after the step in the reference variable to achieve a nearly constant value in just ~5seconds.
- The control variable initially overshoots the target value by ~5% and never becomes completely stable, instead oscillating irregularly about the target value.
- However, the control performance is acceptable for a fast system such as this one.
- The control parameters in this operating mode are nearly ideal for responding to change in the reference variable. This configuration is a compromise between repsonse and control performance.
4.3 Note on Further Experiments
- The experiments and associated parameters described above are a subset of the available possibilities.
- The additiona variants of this demonstration model also easily allow a realization and evaluation of control systems with different settings and control parameters.
Possible Variants:
- Use of just a P-controller
- Use of a PID-controller
-Variations in reference variable step
- Disturbance variable control: Introduction of a disturbance variable z by means of the adjustment cock (6) and comparison of the results with those obtained from reference variable control.
- Optimization of reference and disturbance variable control parameters for various operating points.
Production-related factors, fluctuations in ambient conditions and operational modofications can cause the control system's properties to change.
No comments:
Post a Comment