Home | english  | Impressum | Sitemap | KIT

Evolutionary Algorithm

Optimisation of the Acoustic Resonator Using Evolutionary Algorithms (EA)

Optimising an acoustic resonator is not so easy, because when you tune through one single design parameter, then multiple resonances come and go, amplitudes rise and fall, we want to maximise the sound pressure in the center while avoiding high sound pressures near the walls. This all makes it a tricky optimising task resulting in a smooth, but high-dimensional, multi-modal, non-separable search.

 

We decided to tackle it with evolutionary algorithms, and we decided to build a library dedicated to rapid prototyping of evolutionary algorithms, in order to become able to make educated guesses on suitable EA types to try.

So here is the link to the library:

 

peabox

 

an evolutionary algorithm toolbox written in python