[SOLVED] Analog Circuit optimization using Genetic Algorithm

[Ankitsadat]

Hello Sir/Madam,

I want to use a GENETIC ALGORITHM as an optimization tool for my amplifier design.

I found the many document which describes this all things in theoretical but, I need some practical implementation.

Here I'm using ELDO simulator with mentor graphics ADK_IC cad tool, and RED HAT linux OS
so, kindly help me regarding

1. How to implement a GA for this specific requirement. (How to do and in which language- programming)
2. GA programming related documents.
3. Experimentally developed algorithm with all steps documentation including OS, programming, Simulation, Results etc.
4. online available any kind of package which can directly interface with my simulator.
5. Any other option.

Thank you.

Regards

 

[micpro]

Hello Sir/Madam,

I want to use a GENETIC ALGORITHM as an optimization tool for my amplifier design.

I found the many document which describes this all things in theoretical but, I need some practical implementation.

Here I'm using ELDO simulator with mentor graphics ADK_IC cad tool, and RED HAT linux OS
so, kindly help me regarding

1. How to implement a GA for this specific requirement. (How to do and in which language- programming)
2. GA programming related documents.
3. Experimentally developed algorithm with all steps documentation including OS, programming, Simulation, Results etc.
4. online available any kind of package which can directly interface with my simulator.
5. Any other option.

Thank you.

Regards

Dear Ankitsadat,

are you looking for commercial tools or free tools?
My company (www.muneda.com) sells commercially circuit optimization software. It can read Eldo netlists and there's one module that runs a stochastic optimizer similar to a genetic algorithm.
For amplifier design, I'd rather recommend a sensitivity-based algorithm instead of a genetic one. There will be a large number of structural constraints (saturation, inversion, current symmetries, ...) that restrict the design space very much. Inside this restricted space, the goal function is usually smooth. Genetic algorithms are not efficient for such problems. A sensitivity-based SQP with active set method is usually much faster. We sell that, too.

Regards,
Michael

 

[Ankitsadat]

micpro Sir,

Thanks a lot for your kind suggestion.

Sir, Currently I don't have any idea regarding sensitivity based algorithm. But, I'm interested.

I'm a graduate student so if the commercial version is costly then I can't afford it. is there any evaluation trial version available?
If my institute will help then I will look forward.

Please reply me with detail of pricing and how to order.

Regard.

 

[micpro]

Dear Ankitsadat,

there's no student version, but a cheap university license that your institute could ask for at:
**broken link removed**

Regards,
Michael

 

[Ankitsadat]

Dear Ankitsadat,

there's no student version, but a cheap university license that your institute could ask for at:
**broken link removed**

Regards,
Michael

Thank you Michael Sir.


If any one else have any IDEA then please post here.
Thanks a lot.