Evolutionary VHDL code example

[apsolar]

Hello everyone
Does anyone where I can find a simple VHDl code example based on
evolutionary algorithms.I am doing a project on evolvable hardware.
This will help me get a start on the implementation of Evolvable
Hardware.
Ankit Parikh
Manukau Institute Of Technology

 

[Mike Treseler]

Does anyone where I can find a simple VHDl code example based on
evolutionary algorithms.I am doing a project on evolvable hardware.

VHDL is an excellent hardware
description language, but it is
far from the best text processing language.

You need a custom process for generating
and automatically modifying VHDL synthesis code
based on simulation results.

-- Mike Treseler

 

[apsolar]

Hi Mike
Could you explain your suggestion in a bit detail. I am a novice in
VHDL programming.

Ankit

 

[Mike Treseler]

Could you explain your suggestion in a bit detail. I am a novice in
VHDL programming.

Perhaps you should consider a less complicated project.
My point was that in order to automate the writing
of vhdl synthesis code, you might have to not only
learn vhdl for synthesis and simulation but also
a text processing language such as perl or awk
to handle the automated code modifications.


-- Mike Treseler

 

[Benjamin Todd]

I agree with Mike, VHDL directly may not be the best way to look into a
problem like this.

From what I remember a lot of Genetic Algorithm based work has been done by
modifying FPGA bit files, and using an FPGA that is tolerant of bad
configurations, i.e. doesn't blow-up if you short it by design internally.
(was an old Xilinx device used by Adrian Thompson at Sussex that was capable
of detecting a frequency if I remember rightly)

So you create many bit files, test them all in the apparatus, and give them
a 'fitness' rating. You then mutate and converge the fittest and repeat,
until the bit file you have can do the assigned task.
All very well, but no VHDL involved, I suppose you could fiddle with the
VHDL instead, but here you'll be missing a key point - only the
synthesiseable stuff will make it.
Anyways, this all depends on what level you look at this from, if you're
talking about the feedback and weighting using in machine learning then
you'd have to explain more of your problem.
HTH
Ben

 

[Mike Treseler]

Practical Handbook of Genetic Algorithms:
Complex Coding Systems, Volume III
by Lance D. Chambers
http://tinyurl.com/72ezy