Improved Novel Particle Swarm Optimization Algorithm

Jun Ling
2011-01-01
Computer Engineering and Applications Journal
Abstract:To solve the premature problem of PSO,an improved PSO algorithm with adaptive mutation based on Sobol sequence(SAPSO) is proposed.Based on ICPSO,quasi-random Sobol sequence is introduced to the initialization of the swarm and the adaptive mutation with Beta distribution based on diversity feedback is used to keep the diversity of the population and to avoid the local optimum.The results show the effectiveness of SAPSO solving complicated optimization problems and avoiding the local optimum.The global searching ability is enhanced as well as the convergent speed is guaranteed.
What problem does this paper attempt to address?