 | | Advanced Control strategies
|
|
We are trying to implement advanced control strategies in RCX robots, as neural networks and fuzzy logic. But I think the NQC don't has what we need, as arrays (memory allocation) and aritmetic functions (sigmoid logist for example). Does any one (...) (23 years ago, 5-Apr-01, to lugnet.robotics)
|
| |
| | Re: Advanced Control strategies
|
|
Hi Mauro, I think what you're looking for is legOS. You can find it here : (URL) can't imagine a more appropriate place to ask such questions then this site ! Rainer (...) (23 years ago, 5-Apr-01, to lugnet.robotics)
|
| |
| | Re: Advanced Control strategies
|
|
(...) Lack of floating point support could also be a bit of a problem. You could try legOS. I think you have about 32k available for programs, and there are very few restrictions. You could also do all the hard work on a PC. It's pretty easy to (...) (23 years ago, 6-Apr-01, to lugnet.robotics)
|
| |
| | Re: Advanced Control strategies
|
|
Definately agree on LegOS, my final year Computer Science dissertation was on autonmous learning on an RCX with genetic programming, I used LegOS 2.5, great stuff full C++ if you can do it in ansi you can do it here... But I'd say the best place to (...) (23 years ago, 6-Apr-01, to lugnet.robotics)
|
