[Last update: January 10th, 2017.]
While we have explored heated reactors, surge tanks, heat exchangers and the like, the discussed principles are by no means the only place in which they can be used. In fact, dynamic control and automation are an increasingly relevant (and invisible) part of our everyday lives – the same idea of a feedback/forward loop providing some control over undesired disturbances is relevant and important. Without controllers and feedback, our phones would stay on when we put our face up to them, street lights wouldn’t turn on when it gets dark, and chargers would explode the batteries in our electronics!
The goal of this project is to explore and discuss either an existent form of process control in a normal (i.e. non-engineering) setting, either one that already is used on a regular basis or one that you might like to see implemented in the future. We will spend the first few weeks refining our impression of the process (or processes) on a fundamental, application level – and then later on, use our knowledge of controls to simulate or predict how expected disturbances or perturbations could be managed by our systems.
Last year’s material can be seen here.
Deliverable 0: Group Selection and Project Pitch
Due Date: Friday, February 7th.
As soon as you’ve selected a group of three members, please submit a list of your group members to JLW via email. You may select group members from the other section of the class – however, make sure that all three members are able to attend at least one of the two sections at the same time (8-10AM or 10AM-12PM).
Deliverable 1: Blog Article
Due Date: Working draft by Thursday, March 9th; edited final version by Friday, March 17th.
Each group must write a short (1500-2000 word) blog article on the course site, discussing either a journal article explaining what your novel control process is, or how your selected process can and/or does benefit from process control elements. These articles are designed for an everyday but well-informed audience – think of the regular reader base of Gizmodo or the target demographics of TEDTalks, whom are not primarily focused on the specific vernacular of process control principles and require explanation of these concepts. (Alternatively, think of your target audience as some of your non-engineer classmates, to whom you’re trying to explain the general idea of what you’re doing.) Your article should express the following points:
- What is the application?
- Express the motivation of examining the process or the system that is being considered.
- Why does it need control?
- Establish your controlled variables. What is our basis for determining what is a “good” or “bad” range of operating conditions for this parameter?
- Is the control of this variable critical to its operation, or is it only supportive/additive to its user experience? (For example, a fan that turns on when temperature rises above a certain point might be beneficial in its operation, but isn’t necessary for it to function as a fan – the controls on a Segway, though, not so much.)
- Justification for control should be more substantial than infomercial-style “There’s got to be a better way” statements, even for common household objects – don’t just say it’s an improvement, state why it’s an improvement.
- What about the application needs controlling?
- Identify your manipulated variables. What aspects about your process could you control? Does it make practical sense to do so?
- From this discussion, establish one or two main aspects to use as your points of manipulation to control your controlled variable.
- What could happen to our system?
- Posit potential sources of disturbance that might arise from normal (For example, an automatic hair iron might want to be able to correct for temperature drops when you apply the iron to your hair, but it doesn’t need to account for a sudden current drop from a blackout in your house.)
- How many of these variables could be controlled via the same manipulated variable? Would a feedback or feedforward loop be appropriate for this application? (If you use either of those terms, make sure that you briefly identify what it is for your audience!)
Images, figures, and videos are highly recommended and crucial for effective articles. Consult the WordPress primer on this site or WordPress’s own tutorials for more information. Furthermore, these articles need not be formal in language structure, as they are reflective of your own thoughts and ideas. For this portion of the project, try to be creative and have some fun!
Note that each group must have a functional draft posted to the course website by March 9th that covers all the aforementioned material. However, you may edit your articles to improve them throughout Spring Break.
Deliverable 2: Blog Responses and Retorts
Due Date: Two responses per member, due on the same dates as Problem Sets 4 and 5.
For this portion, students must write two brief (2-3 paragraph) responses to existing posts on the blog, commenting on the validity of the initial article’s application and/or providing additional insight. Note that you’re not anonymous when posting these comments, so please try to keep your opinions substantial and civil. Your response should address the following:
- Does this application make sense? Remark on your impression of the usefulness of the discussed topic. (Positive or negative impressions are both valid.) From a user standpoint, can you envision yourself (or the general public) using this device? If it’s a new application, do you
- Does the proposed control scheme make sense? Based off of the concepts we have discussed in class so far, remark on the selection of the article’s controlled/manipulated/disturbance variables. Are any missing? Would you suggest others? Which would be easier or harder to manipulate?
- Since these applications will require a controller aspect of some type, posit the types of action that might be appropriate for this application. Suggest ways that the developers of a control scheme for this application might reasonably be able to tune/develop controller parameters!
Deliverable 3: Technical Presentation and Simulation
Due Date: Thursday, April 27th.
Equipped with your knowledge of process control gathered over the duration of this course, each group will hold a 20-minute presentation near the end of the semester that builds upon your blog article’s introduction and highlights more technical insight into how your discussed process can be controlled. The audience for this portion will be your peers, so more in-depth information about controller types, stability, etc. will be expected.
Groups will be expected to simulate both plausible perturbation signals on MATLAB, as well as how your process will be able to respond to said signals with your proposed control scheme. Your presentation should feature an approximate model for your blog project process (developed in Simulink) that demonstrates its control of expected perturbations. Some things to consider:
- Posit a reasonable value for time constants for the given processes you are working with. Would the valve/action component of your process operate on the order of seconds? Minutes? Will the process element be of similar time response or significantly slower than your action? Exact numbers aren’t crucial, but orders of magnitude are. Use your intuition of your described process and/or literature values to inform these values!
- Steady state gain values are also necessary, and similarly, you can propose plausible estimates for these values to make sure your streams have reasonable values unto themselves. For instance, you know that an (analog) signal will have a value somewhere between 4-20mA in value; if you are trying to capture temperature fluctuations, for example, make sure that your range of expected temperatures falls within that signal span after the gain is applied! (Seborg Chapter 9 may be useful for this.)
- If you expect that a significant lag time will exist (for a large process, for instance), any reasonable value may be proposed to represent this phenomenon. Again, use literature when possible to justify these numbers or provide a logical justification for why a delay might occur.
- If one of the more advanced methods we’ve discussed in class thus far (i.e. cascade or feedforward control) would be appropriate for your application, it can be constructed in a similar fashion to examples in class, so long as you justify how and why your process can be expressed in this way.