1
$\begingroup$

I'm trying to build a line follower robot and I'm interested in predicting the curves on the track. I have 8 binary sensor array(qre1113). My goal is to make a system that it can generalize what it learned about the curves and give me predictions about where should be at the line to pass it as fast as possible.

How can I integrate a system like Q learning and how can I train it? And also how can I combine this system with a Type C PID controlller ?

There is a paper about it ot you are willing to explain

This is a important project for me and i am kinda running on clock so quick help would be appreciated

$\endgroup$
  • $\begingroup$ You're asking (1) How to implement a neural network, (2) How to train it, and (3) How to use it in conjunction with a PID controller. I've left this question open in the spirit of letting borderline questions stay open, but you have to remember that this is a site of volunteers. That is, someone has to find your question, take time to answer it, and know enough to answer it. If your question were about batteries then you'd have an answer already. Neural networks are an advanced topic; I'm not surprised you don't have an answer yet. $\endgroup$ – Chuck Aug 3 '16 at 18:36
  • $\begingroup$ I personally have asked two questions here - both about the Rigid Body Dynamics Library. The first was open for over a month before the guy that wrote the RBDL answered it! I had a follow-up question I asked the next day, and that was never answered. Point is - when your topics get advanced enough, the pool of people qualified to give answers becomes so small that you're better off contacting an author directly than hoping a question gets answered here, and the paper you link gives contact info. $\endgroup$ – Chuck Aug 3 '16 at 18:40
  • $\begingroup$ Thank you very much for the advice i am updating the link ASAP $\endgroup$ – Ege Keyvan Aug 3 '16 at 18:41

Your Answer

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

Browse other questions tagged or ask your own question.