ELEC 480
– Notes on Grading Proportional Control Projects:
In grading laboratory reports, different aspects of the report are graded
differently. Here are grading schemes for the different aspects of
the report.
Grade for Technical Portion of Report for
Lab
2 (Proportional Control) What is in report
|
Grade
|
What is in the
report.
|
|
0
|
It worked! (Apparently
a miracle occurred.)
|
|
1
|
It worked and
the results checked.
|
|
. . .
|
|
|
6
|
Good experimental
procedure and good data. |
|
7
|
Requirements
for 6 + measured SSE. |
|
8
|
Requirements
for 7 + measured speed of response, possible using two different measures
of speed. |
|
9
|
Requirements
for 8 + predicted SSE and speed of response from open loop transfer function |
|
10
|
Requirements for 9
+ evaluated/compared experimental results with theoretical results and
explained differences. |
You need to be clear
about what needs to be done. Be aware of the following.
-
At this point in the course
you should be able to predict SSE, including any effects on the steady
state produced by offsets, thresholds, etc.
-
You should account for
all offsets, thresholds, etc, and predict the steady state for your system.
-
Then, you should measure
your system's response and compare measurements to theoretical predictions.
-
Explain any discrepancies
to the best of your ability.
-
The essence of this lab
is to be able to merge theory and experiment, and to use that information
to judge (i.e. evaluate) the quality of the control you implement.
-
If you don't do all of
that, you haven't done the lab. The lab is not the measurements and
fiddling with gain values to get something that looks good. (And,
fiddling with two values of gain in a PI controller is much harder than
getting a single gain to work well in a proportional controller.)
Your use of theory to explain measurements is the only way the instructor
can be certain that you understand PID control. The other side of
this is that fiddling with gain values, with no understanding of how to
choose them based on theory will (with high probability) cause you to waste
most of your time, and you may well never get anything that works well.
You will get unexplainable oscillations, for example.