|First Neural Net implemented on a Lego NXT (C like, RobotC)
|Page 2 of 2|
|Author:||Ford Prefect [ Thu Jul 09, 2009 3:43 pm ]|
|Post subject:||Re: First Neural Net implemented on a Lego NXT (C like, ROBOTC)|
aah, you meant the NN topology:
as mentioned, I'd take a Feed Forward Net.
The number of neurons = number of output bits.
1 motor: at least 3 functional states (stop, forward, reverse)
1 neuron may encode 1 bit
10 motors: 3^10 combinations of functional states = 59049 states <= 65536 = 2^16 bits
so you need 16 digital outputs = 16 neurons to encode (control) the functional states of 10 motors
(sry for my poor English)
|Page 2 of 2||All times are UTC - 5 hours [ DST ]|
|Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group