Using deferential evolution algorithm for the multi-objective optimization of a switched-current circuit

The optimization of a circuit based on the switched-current technique is described in this paper. The differential evolution - one of evolutionary algorithms - is used for the optimization. The circuit is applied to low-pass filter realization. The optimization takes into account three requirements. Hence, it is multi-objective (multi-criteria) optimization. The first requirement is the elimination of the effect of nonidealities in the circuit on its magnitude frequency response. The nonidealities are the nonzero output conductance of the field-effect transistors realizing the current sources and the nonzero on-state resistance of the switches. The second and third fact respected by the optimization is concerned with the values of transconductances of the transistors in the circuit. The ratio between the highest and lowest one (the spread of transconductance values) and the sum of all of them should be as small as possible. The reason is a better circuit design. The aim of the optimization is to obtain suitable values of the transconductances satisfying the three requirements.