Particle Swarm Optimizer with Time-Varying Parameters Based on A Novel Operator

R. Cheng,M. Yao
2011-01-01
Applied Mathematics & Information Sciences
Abstract:This paper proposes a time-varying particle swarm optimizer based on our earlier work which introduces a novel operator (leap operator). Two new parameters are recommended in leap operator to prevent premature convergence. With these two parameters, a new modification named LPSO is constructed. Since the values of the 2 parameters are not easy to determine, in this paper, they are modified as time-varying ones. With the time-varying parameters, the modified particle swarm optimizer (TVLPSO) has good potential in finding better solutions. Compared with standard PSO and LPSO, benchmark tests are implemented.
What problem does this paper attempt to address?