I have been looking at papers describing how to do non linear control with neural networks. A lot of these papers have block diagrams with a symbol I don't understand. There is a block with an angled arrow drawn through it. The tail of this arrow is an output of another part of the block diagram.Sample Block Diagram

The sample is from a paper called "Nonlinear System Control Using Neural Networks" by authors Jaroslava Žilková, Jaroslav Timko, and Peter Girovský.

A link to their paper is here:Nonlinear System Control Using Neural Networks

  • $\begingroup$ Should be either shown in the nomenclature or explained in the paper. $\endgroup$
    – Solar Mike
    Apr 24 '19 at 20:20
  • $\begingroup$ Normally in control theory this referring to an adaptive method, so the signal from the angled arrow alters the model of the inverse neural model block. $\endgroup$
    – fibonatic
    Apr 25 '19 at 0:30

Given the title of the figure, I'm inferring that the slanted line means "this block adapts slowly based on the given information". The author is probably trying to convey the notion that you could disconnect that slanted input and things would still work -- for a while.

Block diagrams are of interest to me because they're very much a living language. I work, or have worked, in both control systems and RF signal processing, where there's a lot of overlap between the language, and yet a lot of differences. Watching a controls engineer and an RF engineer show each other block diagrams is like watching someone from London try to talk to someone from Glasgow -- it's officially the same language, but there are significant differences.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.