Evolutionary algorithms based analog filter approximation and implementation

In this work, the denominator coefficients of a lowpass filter transfer function are optimized with evolutionary algorithms in order to obtain minimum approximation error and to reduce the distortion over the passband and stopband separately. For each design case, three different orders of transfer function are optimized. Simulation results show that evolutionary algorithms used in this work results in a short computation time with less approximation error than the conventional methods. Passive and active circuit realizations of filter transfer functions obtained with the most efficient EA method are also provided in order to show the feasibility of the proposed approach for circuit implementation.