Adaptive digital filtering using the bio-inspired firefly algorithm
M. Hussain, W.K. Jenkins · 2017
The Lévy Flights variation of the bio-inspired Firefly Algorithm (LFFA) combines the Firefly Algorithm based on how fireflies interact through flashes and a random-step function based on the Lévy distribution. The LFFA, the Genetic Algorithm (GA), and the Particle Swarm Optimization Algorithm (PSOA) are compared when used for adaptive IIR system identification. Comparisons of the LFFA in direct, cascade, and parallel form structures suggest that 1) the LFFA is more effective than the GA and PSOA, 2) the LFFA adaptive algorithm works well for all structures that were experimentally tested, including a nonlinear Volterra structure, and 3) the LFFA appears to perform most effectively.