Performance of multi-parents genetic algorithms (MPGA) for IIR adaptive system identification

Genetic algorithms (GA) are based on principles of natural selection that originate in biology. The GA has been used for adaptive IIR system identification, but due to slow convergence rates and high computational complexity its use for IIR adaptive systems has been limited. This paper proposes a multi-parents genetic algorithm (MPGA) that is a generalization of the two-parents GA. Results demonstrate that the MPGA can improve convergence rates and maintain relatively low mean-square-errors (MSEs), although it requires increased computational complexity. An attempt to reduce computational complexity is presented and experiments illustrate how the MPGA operates on various digital filter structures.