Title :
Particle swarm optimisation with Kalman correction
Author :
Naha, Arunava ; Deb, Alok Kanti
Author_Institution :
IIT Kharagpur, Kharagpur, India
Abstract :
A novel particle swarm optimisation (PSO) method with guaranteed convergence is proposed which is useful for various optimisation problems. This proposed algorithm searches for the optimum point by the PSO algorithm and at each iteration the optimum location found so far are corrected by the Kalman correction mechanism. This global convergence Kalman PSO (GKPSO) algorithm has been tested for many benchmark problems and the results compared with another popular PSO algorithm with a neighbourhood operator. The proposed algorithm converges faster than the other and also provides better quality of solution. Convergence to the global optimum for this proposed algorithm has been proved.
Keywords :
Kalman filters; convergence; particle swarm optimisation; GKPSO algorithm; Kalman correction mechanism; PSO method; benchmark problems; global convergence Kalman PSO algorithm; global optimum; guaranteed convergence; neighbourhood operator; optimum location; optimum point; particle swarm optimisation;
Journal_Title :
Electronics Letters
DOI :
10.1049/el.2012.4367