Genetic folding/programming toolbox: Analogue circuit design case study

Abstract

Abstract: This work introduces the concept of genetic folding/programming combination to develop a standalone optimisation tool and the developed algorithm is tested with four different benchmark functions. Most GP used in circuit evolution interface two software packages but this work only Matlab is used which reduces the time used for transferring the simulation between the two platforms. To enhance testing of the algorithm and automatic Netlist creation, the expression is extracted with the aid of genetic folding. The automatically simulated Netlist is fed to modified symbolic circuit analysis in Matlab that translate it to matrices to enhance frequency response. The frequency response is then compared to the set frequency response and the RMS difference gives the error which controls the programme towards the desired solution. One circuit is tested and the algorithm successfully evolved the set frequency response.