Assessment#
The grade for this course is determined by the average of two scores: one for the technical report and one for the oral exam. Both scores have equal weight. The report will serve as a basis for the oral examination.
Report#
The results must be presented in a wirtten report. The maximum number of pages is 30, which means you’ll need to be concise. We don’t present you with a definite list of parts that need to be in the report, but you may use the following list as a guide.
Short description of the setup
Explanation of (sensor)calibration (if applicable)
Modeling
Schematic overview of the setup
Definition of symbols (and conventions)
Derivation of equations of motion
System identification / Parameter estimation
Identification strategy
Selection of excitation signal(s)*
Selection of algorithm *
Results *
* Possibly separately if you divide your system into subsystems
Model evaluation (time domain / frequency domain)
Control design
Objectives
Choice of control architectures
Design (x2)
Evaluation - results
Discussion
Writing tips
The main criteria for the report are completeness and clarity. On the basis of the report, any educated reader should be able to reproduce your results with the laboratory setup. All choices made must be properly motivated. Evaluate critically the results obtained. Stick to the 30 page limit!
Content:
Make sure to present sufficient information about the choices you have made and their motivation (for instance, the sampling period, controller design specifications such as the settling time, overshoot, etc., and the identification methods used, etc.)
Modeling: derive the equations of motions. Present the main steps of the derivation, do not oversimplify, the reader must be able to follow the derivation.
System identification: do not forget to report the validation results for your model.
Control: compare simulation results with the results of real-time experiments. Simulate the closed loop using the nonlinear model of the process.
Plot all the important signals (do not forget the control input). Discuss the results, draw conclusions.
Please, do not present theory that can be found in the literature (books, lecture notes, papers), just cite the source. Form:
Follow the usual structure of a technical report (including an Introduction, Conclusions, etc.). Include a Table of Content.
Do not use scanned formulas or tables. Note that typesetting a formula in Word or LaTeX typically takes less time than scanning it.
Do not use the “Matlab” …e-5 notation, use …x10^-5 instead (\times or \cdot in LaTeX).
Terminology: use established technical terms and symbols as much as possible.
Explain the meaning of all symbols - mathematical symbols must be explained at their first use and your document should preferably include a list of symbols.
When comparing two or more signals (e.g., model output and data) display them in one plot, using different line types (not only colors, your reports can be printed in black and white). Explain the colors / line types. The scale of the plots must be carefully chosen in order to clearly convey the information intended.
Label properly the axes in graphs (variables and units). Replace the standard Matlab label “Amplitude” by the respective quantity displayed. Make sure that the axes labels are not too small.
In the text and figure labels or captions, units are typeset in Times Roman (not italics), functions like sin, cos, etc. are typeset in Times Roman, not italics (in LaTeX, use \sin, \cos, etc.)
Figure captions are below the figures, table captions above the tables.
Block diagrams must be explained (the function of the blocks).
Statements like “the prediction-error method is used” are not informative, there are many such methods, be more specific.
Strive for elegant, concise text - longer reports do not necessarily yield higher grades.
Number the pages and check for broken or undefined references (?? in LaTeX).
English language:
Do not use short forms, like “isn’t”, “wouldn’t”.
Do not use colloquial style, like “a couple of”.
Spell check and proofread your report.
The use of initial capitals in chapter and section titles must be uniform throughout the report.
When referring to sections, figures, etc., capitalize, like: see Figure 1.
Some typical mistakes:
Words like openloop model are written with a dash: open-loop model
“More easy”, “more simple”, etc. should be “easier”, “simpler”, etc.
There are no spaces before commas, like in the above phrase.
However, there is a space before an opening parenthesis (…), not like this(…).
The report must be delivered on paper - two copies in a dedicated box, which will be placed in the vicinity of the mailboxes on the first floor, next to the DCSC secretariat. Please, do not bind the report, just staple it with one staple in the upper left corner. This will make it easier for us to handle and read.
Along with the two hardcopies of the report, upload a copy of your report and the (Matlab)-files required to run your experiments in a zip-file to Brightspace (under Assignments
)
Exam#
The (oral) exam will be held outside of the basement lab, with the group members and two staff members present. The exam itself will take 30 minutes. We’ll pick you up at the lab, make sure that you’re logged into the PC and have a demonstration ready for showing after the discussion.
The staff members will base the questions on the content (or sometimes lack of) content in your report. We expect both team members to equally contribute to the discussion, and both team members to at least have knowlede on all aspects of the project. It is however acceptable if more advanced / detailed questions are answered by a specific team member (as long as the other member has some other area of expertise). At the end of the discussion, the staff members will evaluate the discussion in private, and the team members go down to the lab to watch the demonstration that has been prepared. After the demo, we’ll provide you with feedback, and we strive to directly give you an indication of your final grade.