Genetic algorithm based adaptive system identification


KARABOĞA N. , CETINKAYA B.

IEEE 15th Signal Processing and Communications Applications Conference, Eskişehir, Turkey, 11 - 13 June 2007, pp.145-148 identifier identifier

  • Publication Type: Conference Paper / Full Text
  • City: Eskişehir
  • Country: Turkey
  • Page Numbers: pp.145-148

Abstract

In this paper, the application of adaptive system identification based on genetic algorithm is realized. A high order unknown plant is modeled with a lower order adaptive Finite Impuls Response (FIR) filter. Also, a performance comparison has been made between the conventional gradient based Least Mean Square (LMS) algorithm and genetic algorithm (GA). Simulation results show that the genetic algorithm based design method has a better performance in terms of convergence speed and error performance.